delivered the opinion of the court: The defendant, Derrick J. Wheeler, was convicted of first degree murder (720 ILCS 5/9 — 1(a)(1) (West 1996)) in the Will County circuit court. He claims that his convictions should be reversed because (1) he was not proven guilty beyond a reasonable doubt, and (2) the court erroneously admitted evidence from “invisible blood” testing. We remand for a Frye hearing on the “invisible blood” testing procedure.
BACKGROUND
Before the defendant’s trial, his attorney filed a motion in limine seeking, inter alia, to bar the State from presenting evidence from a Leuco-Malachite Green (LMG) test the police performed on his car. LMG is a chemical mixture used to detect latent blood on a surface. The defendant’s attorney argued that the test results were highly prejudicial and lacked probative value. The judge reserved his ruling on the LMG portion of the motion.
The case proceeded to trial on September 9, 1998. The State’s first witness was Jacques Buckley, who testified that he was a lifelong friend of Monte Love, the victim. Buckley said he and the defendant were “riding around” in the defendant’s yellow car in the early morning hours of June 18, 1997. They eventually picked up Love and drove to the defendant’s house, where Buckley and Love entered and bagged some “fake dope.” They exited the house after about 15 minutes, and the defendant said, “Let’s go.” Karen Washington and another person *276were also outside the house at that time. The three men left in the defendant’s car with Buckley driving, Love in the front passenger seat, and the defendant in the backseat.
According to Buckley, they drove around the neighborhood and eventually detoured from Patterson Road because a train was crossing in front of them. Love then asked Buckley to stop the car so he could get out and relieve himself. Buckley said he stopped under a noisy highway, and Love exited the car followed by the defendant. Buckley then heard two quick gunshots. He looked back and saw Love lying on the ground and the defendant returning to the car. He began pulling away, and the defendant entered through the front passenger door with a gun in his hand. Buckley asked what happened, and the defendant told him not to worry about it. The defendant placed the gun in a bag and put it under the seat.
Buckley testified that he saw a white woman with a dog or a cat as they left the scene. He said they drove to a store, where he used the bathroom and left in the car with another man, named Lamaris George. The defendant was not with them at that time. Buckley said he told a man named Jermaine Holifield about the shooting the day after it happened. He acknowledged that he did not speak to the police about the shooting until a detective contacted him approximately 10 weeks later while he was in jail for a pending drug case. He testified that the defendant had threatened to “do” him also if he told anyone about the shooting.
In addition to this testimony, Buckley acknowledged that he had been convicted of three felonies and a misdemeanor, all involving acts of dishonesty. He also acknowledged that he was involved in a pending case for narcotics possession. In that case, in exchange for his testimony against the defendant, the State agreed that he would receive 30 months of probation, 180 days in the county jail, and drug counseling. The State also granted him immunity from any murder charges stemming from Love’s death.
The parties stipulated that Karen Washington would offer the following testimony if called as a witness. She was at the defendant’s house between 6:30 and 7 a.m. on the morning of the shooting. She observed Buckley, Love, and the defendant arrive in a yellow Oldsmobile Delta 88. The top of the Oldsmobile was the same color as the bottom, and she believed it belonged to the defendant. The three men left in the Oldsmobile around 7 a.m., after the defendant mentioned having something important to do. Buckley was driving, the defendant was in the front passenger seat, and Love was in the backseat. Washington heard that Love was dead about 30 to 45 minutes later.
Reginald Pinnick, one of Love’s friends, testified that he saw Love *277riding in the front passenger seat of a yellow car around 7 a.m. on the day of the shooting. He tried to stop the vehicle to speak with Love, but Love raised one finger as if to indicate he would be right back. Pinnick saw two other persons in the car whom he could not identify. He said he could not recall whether the car had a vinyl top and did not remember telhng police detectives it had a vinyl top. He also said the car resembled a yellow Oldsmobile Delta 88 depicted in People’s exhibit No. 6 (a photograph), but he could not definitely identify it as the Delta 88.
The parties next stipulated that a train crossed Patterson Road at approximately 7:35 a.m. on the morning of the shooting.
Camille Sharp testified that Interstate 80 is elevated “over the top of [her] backyard.” She said on the morning of the shooting, she went outside to get her dog and saw a man urinating beside a car that was facing her house. After unchaining her dog, she looked up and saw a second man, with a gun, standing behind the man who was urinating. The second man initially made an unsuccessful attempt to fire the gun. However, he made another attempt and shot the first man in the back of the head. Sharp saw “a big cloud of blood” followed by the victim falling backward to the ground. She said the shooter’s arms were outstretched in opposite directions, one reaching for the car door and the other pointing the gun. She further said the distance between the victim and the car was a “little over” the length of the shooter’s outstretched arms.
According to Sharp, the shooter fired two additional shots at the victim while attempting to enter the car. He then entered the car and it pulled away. Sharp believed there were two persons in the car at that time. She described the car as a large, white, four-door vehicle with chrome bumpers and a vinyl top that was a shade darker than the body.
Sharp called 911 after witnessing the shooting. She testified that she observed the shooter’s complexion and described his clothing to a police detective. When asked about the information in the detective’s report, she remembered telhng him the shooter wore a sports jersey with black and blue lettering on it and had a black “do-rag” on his head. However, she said she did not recall telling him the shooter’s complexion resembled a dark Mexican or a fight black man. Instead, she testified that she believed the shooter’s skin was darker than the victim’s skin.
Sharp viewed three or four cars at the police station and identified the one that most resembled the car used in the shooting. However, she later called the police and directed them to another car she had seen on her own. She testified that she directed them to the other car to show them what the car used in the shooting looked like.
*278Police officer James Kren testified that he arrived at the scene at approximately 7:30 a.m. He found a dead black male lying face-up under Interstate 80 with blood covering the right side of his face.
Louis Bolognani, a police evidence technician, testified that he photographed the murder scene and collected evidence. Upon arriving at the scene, he found the victim lying face up with the zipper of his trousers down. Bolognani preserved casts of tire impressions made by a vehicle that apparently left the area at “an excessive rate of speed.” He discovered two small plastic bags containing a white powdery substance, one on the victim’s chest and another by his left hand. Bo-lognani also took tire impressions and lifted latent fingerprints from three vehicles in police custody, including the yellow Oldsmobile depicted in People’s exhibit No. 6. He did not testify regarding any connection between the defendant and the impressions or the fingerprints. He said he found no evidence inside the yellow Oldsmobile, describing its interior as “exceptionally clean.” He also said the Oldsmobile did not have a vinyl top.
The parties stipulated that neither of the plastic bags recovered from the victim contained a controlled substance and that the victim’s death resulted from gunshot wounds to the head.
The State’s next witness was William Smith, an evidence technician who performed an LMG test on the defendant’s car. Before Smith took the stand, the defendant’s attorney renewed his argument that the LMG evidence should be barred because it was highly prejudicial and lacked probative value. He further stated:
“I believe the State if they wanted to present evidence of this, I think there is a Frye test. We talked about this earlier, as to whether this is a test that’s generally accepted by the scientific community for the purposes of which it’s being used to show this is an indicator of blood or not.”
The judge subsequently denied the portion of the defendant’s motion in limine relating to the LMG evidence.
Smith then took the stand and testified that he had been a police evidence technician for approximately nine years. He had received evidence training at Northwestern University, including instruction on the use of LMG as a blood-enhancement technique. He explained that the LMG test involves spraying a chemical mixture on a surface and watching for a color change to “dark blue green.” Such a change, he stated, “will give you a presumption to say that this is blood.” He acknowledged that possibly as many as 50 other substances, including fresh potato juice and certain metals, would cause the same reaction. When discussing that fact, he stated: “It’s presumed to be blood. It could be another substance, but it is presumed to be blood.”
*279Smith then testified that he performed an LMG test on the defendant’s yellow Oldsmobile Delta 88 during the murder investigation. He did not see any blood on the car beforehand. He said he sprayed the inside of the rear passenger-side door and observed “a couple little dots” where the LMG exhibited the dark blue color change.
Detective Steve Bajt was the State’s final witness. He testified that he tried to speak with Lamaris George to verify Buckley’s claim that they were together after the shooting. However, George’s parents would not let Bajt see him, stating that he could not communicate because of a medical condition. Bajt never spoke with Jermaine Holi-field.
Bajt testified that he arrested the defendant and took the Oldsmobile Delta 88 into custody. The defendant waived his Miranda rights and denied knowing Love or being involved with the shooting. However, after looking at a photograph, he said he recognized Love as “Demonstration Man” and acknowledged giving him a ride to the projects on the day of the shooting.
According to Bajt, the defendant gave a statement containing the following facts: Love came to the defendant’s house on the morning of the shooting wanting to buy some “weed.” Buckley was outside at the time with some friends, including Karen Washington. The defendant advised him that he did not have any “weed,” and Love asked for a ride to the projects. Buckley and the defendant took him to the projects in the defendant’s car. Buckley was driving, the defendant was in the front passenger seat, and Love was in the backseat. The defendant did not drive because his license was suspended. After dropping Love off at the projects, Buckley and the defendant drove to see a girlfriend but discovered that she was not home. The defendant’s car would not start when they tried to leave, so Buckley got a ride with another person to pick up a battery.
Bajt testified that, while going over the statement a second time, the defendant said he went with Buckley to pick up the battery and they both returned and fixed the car. He never identified the person who took them to get the battery. Bajt also testified that the Oldsmobile Delta 88 he took into custody did not have a vinyl roof.
The State then rested, and the defendant recalled Bajt to testify. Bajt said he interviewed Buckley in jail and advised that his name had surfaced during the investigation of Love’s murder. Buckley never mentioned seeing a white woman with a dog, telling Jermaine Holi-field about the shooting, or being threatened by the defendant. Bajt also interviewed Reginald Pinnick, who described the car he saw as a beige or yellow four-door vehicle with a vinyl top. Additionally, Bajt *280interviewed Camille Sharp at the scene of the shooting and at the police station. Sharp described the shooter as a male in his mid-twenties, either a dark Mexican or a light-complected black man, wearing a black sport jersey with blue lettering and a black “do-rag” on his head.
Three stipulations were presented after Bajt’s testimony. The first stipulation indicated that the tire impressions at the scene of the shooting did not match the tires on the defendant’s car or any other cars in police custody. The second stipulation indicated that during his grand jury testimony, Buckley never testified that he (1) saw a white woman with a dog, (2) told Jermaine Holifield about the shooting, or (3) was threatened by the defendant. The third stipulation indicated that if Buckley was recalled to testify, he would acknowledge being addicted to drugs dating back to the day of the shooting.
After hearing the evidence, the jury found the defendant guilty of first degree murder. He filed a motion for a new trial claiming, inter alia, that the judge erred in admitting Smith’s testimony about the LMG test without conducting a Frye hearing. The motion was denied, and the defendant was sentenced to 48 years in the Department of Corrections. He now appeals.
ANALYSIS
I. Reasonable Doubt
The defendant first claims that the State’s evidence was insufficient to prove him guilty beyond a reasonable doubt.
It is well established that a jury’s role is to determine the credibility of the witnesses, weigh their testimony, and draw inferences from the evidence. People v. Jimerson, 127 Ill. 2d 12, 535 N.E.2d 889 (1989). An appellate court’s role, when addressing a challenge to the sufficiency of the State’s evidence, is to determine whether any rational trier of fact could have found the elements of the crime proven beyond a reasonable doubt. People v. Collins, 106 Ill. 2d 237, 478 N.E.2d 267 (1985). In making this determination, the court considers the evidence in a light most favorable to the State. Collins, 106 Ill. 2d 237, 478 N.E.2d 267. “Where evidence is conflicting, it is the prerogative of the jury to ascertain the truth; a reviewing court may not substitute a different conclusion on questions regarding the credibility of witnesses.” People v. Gill, 169 Ill. App. 3d 1049, 1053, 523 N.E.2d 1239, 1242 (1988). A jury verdict should not be disturbed unless the evidence is so improbable, unsatisfactory, or inconclusive that no rational trier of fact could have found the defendant guilty beyond a reasonable doubt. People v. Jarvis, 158 Ill. App. 3d 415, 511 N.E.2d 813 (1987).
*281In light of these principles, we conclude that the State presented sufficient evidence to sustain the defendant’s conviction for first degree murder. Jacques Buckley testified that he witnessed the murder and that the defendant was the shooter. His testimony was corroborated in several respects by other witnesses. For example, Karen Washington confirmed that the defendant left his house on the morning of the murder with Buckley and the victim. She also confirmed that they left in the defendant’s car and that Buckley was driving. Reginald Pinnick testified that he saw the victim riding in the front passenger seat of a yellow car with two other individuals that morning. The parties stipulated that a train crossed Patterson Road at the time Buckley claimed to have been detoured by one. Officer Bolognani testified that the victim was found dead with his trousers unzipped, corroborating Buckley’s statement that the victim was shot after exiting the car to relieve himself.
Additionally, Camille Sharp witnessed the shooting and testified that she saw two men standing beside a car near an elevated portion of Interstate 80. She said one of the men was urinating while another man shot him in the back of the head. She said the shooter then entered the car and it pulled away. She believed two persons were in the car at that time. These details corroborate Buckley’s description of the shooting.
We acknowledge the defendant’s argument that Buckley was not credible. However, Buckley’s credibility was a matter for the jury to determine. After viewing the evidence in a light most favorable to the State, we cannot say that no rational trier of fact could have found the defendant guilty beyond a reasonable doubt.
II. “Invisible Blood” Evidence
The defendant next claims that Smith’s testimony about the LMG test should not have been admitted without a Frye hearing, and that the testimony was irrelevant.
The State contends that the defendant has waived his Frye claim by failing to assert it in the trial court. However, the record shows that the defendant’s attorney mentioned the Frye test while arguing his motion in limine. The attorney also alluded to previous discussions before the court regarding application of the test. Furthermore, the defendant’s motion for a new trial contained a claim that the judge should have conducted a Frye hearing. We decline to apply the waiver doctrine under such circumstances. See Harris v. Cropmate Co., 302 Ill. App. 3d 364, 706 N.E.2d 55 (1999) (declining to apply the waiver doctrine even though the defendant failed to request a Frye hearing in the trial court).
*282When determining whether an expert witness can give an opinion based on a scientific theory, Illinois courts follow the test set forth in Frye v. United States, 293 F. 1013 (D.C. Cir. 1923). The Frye court explained the test as follows:
“Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.” (Emphasis added.) Frye, 293 F. at 1014.
Thus, expert testimony based on a scientific theory is not admissible unless the theory has gained general acceptance in the expert’s field. People v. Ferguson, 172 Ill. App. 3d 1, 526 N.E.2d 525 (1988). The determination of whether a theory has gained such acceptance is made at a separate evidentiary hearing. See Harris, 302 Ill. App. 3d 364, 706 N.E.2d 55.
These principles apply to determining the admissibility of Smith’s testimony. The prosecutor argued otherwise in the trial court, asserting that Smith would not testify as a scientific expert because he “would not *** give an opinion as to whether or not the substance which was sprayed was actually blood.” Rather, according to the prosecutor, Smith would simply “testify to the test that he performed and what the results were.”
These statements understate what Smith did when he took the stand. He described the LMG test and the reaction he observed on the door of the defendant’s car. However, he further told the jurors that a reaction with LMG raises a presumption that the detected substance is blood, even though possibly as many as 50 other substances could cause the same reaction. He gave this testimony against the backdrop of his training at Northwestern University and his years of experience as an evidence technician. His testimony amounted to an opinion that the reaction on the car door was a basis for presuming the presence of blood over all the other possible substances. Such an opinion falls within the realm of expert testimony.
We are convinced that a remand for a Frye hearing is warranted. The degree to which LMG testing is accepted in the scientific community is unclear. No reported cases in Illinois address the admissibility of LMG evidence. Additionally, the trial judge’s comments on the record suggest that he was unfamiliar with LMG. The State cites People v. Henne, 165 Ill. App. 3d 315, 518 N.E.2d 1276 (1988), and *283 People v. Hendricks, 145 Ill. App. 3d 71, 495 N.E.2d 85 (1986), for the proposition that Luminol testing, which is similar to LMG testing, involves a generally accepted procedure. We note that Hendricks has been reversed (People v. Hendricks, 137 Ill. 2d 31, 560 N.E.2d 611 (1990)). In any event, neither case is helpful because they merely mention the use of Luminol without addressing the admissibility of the test results.
We realize that the LMG testing procedure is taught at Northwestern University. However, we are not willing to assume that the Frye standard is satisfied based solely on that fact. Furthermore, the issue in the instant case involves more than just the testing procedure. In light of Smith’s testimony, the relevant inquiry is whether LMG testing is generally accepted in the scientific community as a means of raising a presumption (rather than just a possibility) that blood is present. That inquiry should be addressed at the Frye hearing on remand. Unless it is determined that Smith’s testimony qualifies for admission under the Frye standard, the defendant should receive a new trial, because we do not believe the resulting error would be harmless. See Ferguson, 172 Ill. App. 3d 1, 526 N.E.2d 525 (noting that a retrial under such circumstances does not subject a defendant to double jeopardy).
The defendant also claims that Smith’s testimony should not have been admitted because it was irrelevant. Since that claim would remain if the testimony passed the Frye standard on remand, we will address it now. Evidence is relevant if it has any tendency to make a fact in consequence to the determination of the action more or less probable than it would be without the evidence. People v. Hope, 168 Ill. 2d 1, 658 N.E.2d 391 (1995).
The State presented evidence that the defendant exited the rear passenger-side door of the Delta 88 and shot the victim near the car, causing a large cloud of blood. Accordingly, the existence of blood on the door was a fact of consequence to the determination of the case. Smith’s testimony made that fact more probable and thus qualifies as relevant. For these reasons, we cannot conclude that the judge abused his discretion in admitting Smith’s testimony as relevant.
CONCLUSION
The cause is remanded to the circuit court for further proceedings consistent with this order. This court retains jurisdiction of the appeal pending resolution of the Frye issue in the circuit court.
Remanded with directions.
SLATER and LYTTON, JJ., concur.
*284SUPPLEMENTAL OPINION
JUSTICE HOMER delivered the supplemental opinion of the court:
Defendant Derrick J. Wheeler was convicted of first degree murder (720 ILCS 5/9 — 1(a)(1) (West 1996)). He appealed, challenging the sufficiency of the State’s evidence and the admission of testimony of “invisible blood” testing. As previously set forth, this court found the State’s evidence sufficient to sustain defendant’s conviction, but we retained jurisdiction and remanded the cause to the trial court for a hearing to determine the degree to which Leuco-Malachite Green (LMG) testing is accepted in the scientific community.
After a hearing on January 15, 2002, the trial court ruled that LMG evidence was properly admitted. We review Frye issues under an abuse of discretion standard. Donaldson v. Central Illinois Public Service Co., 199 Ill. 2d 63, 767 N.E.2d 314 (2002). Based on our careful review of the record, as supplemented by a transcript of the hearing and supplemental briefs of the parties, we now reverse defendant’s conviction and remand the cause for a new trial.
A detailed statement of the facts of the case is set forth above. We will include only those facts here that are relevant to the LMG testing procedure.
Prior to trial, defendant filed a motion in limine to bar the State from presenting testimony by Officer William Smith relating to LMG testing of the door of defendant’s car. Defendant argued that the testimony was unduly prejudicial and that the State had not established that the test was generally accepted in the scientific community as an indicator of blood. Upon the prosecutor’s assurance that Smith would not testify as an expert witness, the trial court admitted the testimony.
Smith then testified that prior to spraying a solution of LMG on the car door, he did not observe any spots or substances resembling blood. The spray revealed a “couple” of spots that gave a positive reaction by turning dark blue-green. Smith stated that LMG is a “presumptive test” for the presence of blood and that a positive LMG reaction “will give you a presumption to say that this is blood.”
Because Smith’s testimony amounted to an expert opinion, this court remanded the cause for a hearing pursuant to Frye v. United States, 293 F 1013 (D.C. Cir. 1923). We noted that no reported cases in Illinois had addressed the admissibility of LMG evidence and that the issue in this case required more than simply determining whether LMG was a testing procedure generally accepted in the scientific community. We directed the court to determine specifically “whether LMG testing is generally accepted in the scientific community as a means of *285raising a presumption (rather than just a possibility) that blood is present.” 334 Ill. App. 3d at 283.
At the hearing on remand, two forensic scientists testified, Dr. Robert Gaensslen for the State and Marc Scott Taylor for the defense. Dr. Gaensslen testified that LMG testing is generally accepted in the scientific community as a “presumptive” or “preliminary” test that produces a positive reaction to substances containing a sufficient amount of peroxidase enzyme. He explained that approximately 50 different substances — including, for example, copper and blood — contain peroxidase enzyme and give a positive reaction to LMG. On cross-examination, however, Gaensslen said that if a suspected substance such as blood was not visible to the naked eye, a positive LMG test indicated a “possibility” of its presence, but a “confirmatory” test would have to be performed to determine whether the substance was blood.
Defendant’s expert, Taylor, agreed with Dr. Gaensslen that a positive LMG reaction on an invisible stain raised only a possibility, but not a presumption, that blood was present. He said a stain was presumed to be blood when it was visible and looked like blood and then produced a positive reaction to LMG.
The Frye hearing established that a positive LMG reaction indicates a broad range of possible substances present on the material tested, but it does not, without more, permit the tester to “presume” the presence of any particular substance within the category of matter containing peroxidase enzymes. The experts agreed that if a suspected substance is not visible, LMG testing alone could only indicate the possible presence of blood. In other words, although LMG is generally accepted in the scientific community as a preliminary test for the presence of blood, it is not an accepted method for presuming the presence of blood as Smith opined at trial.
Thus we conclude that, at least in the absence of visible evidence of blood, LMG testing is not generally accepted in the scientific community as a means of raising a presumption (rather than just a possibility) that blood is present.
Our conclusion today is generally consistent with decisions in sister jurisdictions where the use of “presumptive” tests has been challenged. See, e.g., Brenk v. Arkansas, 311 Ark. 579, 847 S.W.2d 1 (1993) (presumptive test improperly admitted in absence of confirmatory test to determine whether substance causing reaction was blood); State v. Fukusaku, 85 Haw. 462, 946 P.2d 32 (1997) (presumptive testing properly excluded because, without confirmatory tests, prejudicial effect of evidence was not outweighed by its probative value); State v. Canaan, 265 Kan. 835, 964 P.2d 681 (1998) (evidence of *286presumptive tests admitted where confirmatory testing established that substance causing reaction was blood); State v. Peterson, 242 Neb. 286, 494 N.W.2d 551 (1993) (LMG evidence properly admitted where visible dark stains indicated presence of blood and confirmatory tests established that substance was blood); cf. Commonwealth v. Duguay, 430 Mass. 397, 720 N.E.2d 458 (1999) (evidence properly admitted where expert testified that presumptive ori/m-tolidine test used to detect blood was a screening test capable of producing false positives).
In sum, we hold that the testing procedures utilized in the instant case did not give rise to a presumption as to the presence of blood on the door of defendant’s car. Therefore, the admission of Smith’s expert opinion to the contrary resulted in reversible error. Accordingly, we reverse defendant’s conviction and remand the cause to the circuit court of Will County for a new trial.
Reversed and remanded.
LYTTON, EJ., and SLATER, J., concur.