Appellant, Tytus Antwion Houston, appeals the judgment of the Pulaski County Circuit Court convicting him of attempted first-degree murder, burglary, and misdemeanor theft of property. Appellant was tried by a jury and sentenced to consecutive terms of eighteen years for the attempted murder and fifteen years for the burglary, with the misdemeanor merging into the felonies, resulting in a cumulative sentence of thirty-three years imprisonment. Jurisdiction of this appeal is properly in this court pursuant to Ark. Sup. Ct. R. l-2(a)(2). The *600sole point of error urged for reversal is the exclusion at trial of luminol test results. We find no error and affirm the judgment.
The evidence produced at appellant’s trial showed that appellant entered a residence through a window and attacked the victim while she was sleeping. Appellant beat the victim with his fists, stabbed her with a kitchen knife, and beat her with a hammer. There was testimony that the victim was covered with a large amount of blood and that blood was in her bedroom and in other rooms of her house. In anticipation of this blood evidence, appellant sought to introduce evidence that a luminol test for the presence of blood on the leather jacket he purportedly wore during the attack yielded negative results. The sole issue presented in this appeal is whether the trial court abused its discretion in denying the admission of that evidence.
Our law is well-settled that luminol test results indicating the presence of blood are inadmissible without follow-up tests confirming the presence of human blood related to the crime. Young v. State, 316 Ark. 225, 871 S.W.2d 373 (1994); Palmer v. State, 315 Ark. 696, 870 S.W.2d 385 (1994); Brenk v. State, 311 Ark. 579, 847 S.W.2d 1 (1993). Because luminol testing can return false positive results by reacting with substances other than human blood, and because luminol testing is not time-specific, luminol test results are not relevant per se and their admission without additional factors that relate that evidence to the crime would confuse a jury. Id.; accord Young, 316 Ark. 225, 871 S.W.2d 373; Palmer, 315 Ark. 696, 870 S.W.2d 385.
In announcing these rules on the admissibility of positive luminol test results, this court applied the relevancy approach to the admission of novel scientific evidence enunciated in Prater v. State, 307 Ark. 180, 820 S.W.2d 429 (1991). Brenk, 311 Ark. 579, 847 S.W.2d 1. Under that approach, the trial court must conduct a preliminary inquiry to determine the reliability of the novel process used to generate the evidence, the possibility of confusing or misleading the jury if the evidence is admitted, and the connection between the proffered evidence and the disputed factual issues in the particular case. Id., citing Prater, 307 Ark. 180, 820 S.W.2d 429.
In the present case, the trial court heard, in chambers, the testimony of Stan Wilhite, the member of the crime scene inves*601tigation unit of the Little Rock Police Department who conducted the luminol test on the leather jacket. Mr. Wilhite testified that luminol tests produce positive results for many substances other than human blood. When the trial court effectively asked if luminol discloses negative results in the presence of blood, Mr. Wilhite responded that he could not answer the question. The trial judge then ruled that, given the focus of the Brenk and Palmer cases on the possibility of false positive results, he could not allow the evidence since he had no testimony indicating luminol tests did not also produce false negative results.
On the record before us, we cannot say the trial court abused its discretion in so ruling. There was no evidence of the reliability of the luminol test for proving the absence of human blood. Reliability is the critical element in the relevancy approach to admissibility of novel scientific evidence. Prater, 307 Ark. at 186, 820 S.W.2d at 431. The proponent of the novel scientific evidence carries the burden of proof on the reliability issue. See id.; cf. Rock v. State, 288 Ark. 566, 708 S.W.2d 78 (1986), vacated on other grounds, 483 U.S. 44 (1987) (staling that proponent of testimony by witness who has undergone hypnosis bears burden of proving reliability). It follows that, as the proponent of the negative luminol test results, appellant bore the burden of demonstrating the reliability of the luminol test for showing the absence of human blood. It is clear from this record that appellant failed to carry that burden, as Mr. Wilhite could not testify to the accuracy of luminol tests for proving the absence of human blood. Therefore, we cannot say the trial court abused its discretion in denying the admission of the negative luminol test results.
The judgment is affirmed.