delivered the opinion of the court:
The defendant, Charles J. Acri, was charged with one count of arson (720 ILCS 5/20 — 1(b) (West 1992)) and one count of insurance fraud (720 ILCS 5/46 — 1 (West 1994)) after allegedly setting fire to his home in East Moline, Illinois. In response to the defendant’s motion in limine, the trial court barred the State from introducing evidence that a dog trained to detect the presence of accelerants at a fire scene had alerted at various places in the remains of the home. The State appeals. We affirm.
On July 18, 1994, the defendant’s home was damaged by fire. Two days after the fire, a special agent for the State Fire Marshall’s office went to the scene of the fire with a dog specially trained to assist in fire scene investigations by detecting the presence of acceler-ants. The dog "alerted,” or indicated that it had detected the presence of an accelerant, in the carpet along three walls of the screened-in porch and at an area of "spalling” in the center of the concrete floor of the porch. (Spalling is the distinctive mark made when concrete is exposed to extreme heat and the moisture in the concrete turns into steam.) The agent did not take samples from the areas where the dog had alerted because he knew that samples had been taken previously and because the area had been unsecured for two days. Laboratory analysis of those samples previously taken was *1032negative for the presence of accelerants. The record does not indicate whether these samples were taken from the areas where the dog alerted.
Prior to trial, the defendant filed a motion in limine in which he sought to exclude, inter alia, evidence that the arson investigation dog had alerted to the presence of accelerants in the remains of the defendant’s home and evidence that would tend to show that dogs can smell the presence of accelerants in amounts too small to be detected by laboratory analysis. The defendant argued that this kind of evidence had not gained general acceptance in the field of arson investigation.
At the hearing on the defendant’s motion, the State called William Glover, a special agent with the State Fire Marshall’s office and the handler of the dog involved in this case. Glover testified in great detail about the training of his dog, Watson, and Watson’s ability to detect accelerants present at various fire scenes. He also testified about Watson’s alerts in the remains of the defendant’s home. Glover also described a study conducted by a chemistry professor at Illinois State University in which Watson was able to detect the presence of accelerants even though the chemical analysis of the samples was negative. Glover told the story of a fire in Monmouth, Illinois, in which Watson had detected the presence of an accelerant. Laboratory tests for accelerants were negative, but Watson’s "alerts” were confirmed by the confession of the arsonist.
Glover indicated that he was a member of the International Association of Arson Investigators (IAAI). He was aware of the position paper published by the IAAI which indicates that evidence of a canine alert should be used in criminal cases only when laboratory analysis supports the canine’s findings. Glover stated that he did not agree with that position. Moreover, he added that the Canine Ac-celerant Detection Association (CADA), of which he is a member, does not agree with the IAAI’s statement and that several letters had been written by people in the arson investigation field which opposed the position of the IAAI.
John Marcouiller, a chemist with the Illinois State Police Forensic Science Laboratory, testified that he analyzed the samples taken from the defendant’s home. He agreed with the IAAI’s position that evidence of an alert by a dog should not be used without laboratory findings to confirm the presence of accelerants. In his opinion, dogs are a tool that should be used to determine whether an arson has taken place, but they should not be the only tool. He thought that most of the other arson examiners in the State would agree with him on that point.
*1033Another witness to testify at the hearing was Mike Hiles of the Iowa State Fire Marshall’s office. Hiles is a dog handler whose job is similar to that of William Glover. He indicated that he did not agree with the IAAI’s position on the use of accelerant-detection canines. In his opinion, the people who wrote the policy were not dog handlers and were unfamiliar with a dog’s ability to detect accelerants in amounts too small to register on laboratory tests.
Finally, the parties admitted into evidence the IAAI position paper, the CADA response and the article which detailed the study conducted at Illinois State University using William Glover and his dog.
The trial court granted the defendant’s motion in limine in part and ruled that the State could not offer testimony regarding the alerts made by Watson at the defendant’s home. The court held that this area of expertise had not gained general acceptance by the scientific community.
On appeal, the State argues that the trial court abused its discretion in barring the evidence of the canine accelerant-detection unit even though the dog’s alerts were not confirmed by laboratory analysis.
The admission of scientific evidence in Illinois is governed by the test set forth in Frye v. United States (D.C. Cir. 1923), 293 F. 1013. (People v. Eyler (1989), 133 Ill. 2d 173, 549 N.E.2d 268.) Under the Frye test, evidence may be admitted when the scientific principle on which it rests has gained general acceptance in its particular field. (Frye, 293 F. 1013.) The determination of whether to admit scientific evidence lies in the sound discretion of the trial court. People v. Mack (1989), 128 Ill. 2d 231, 538 N.E.2d 1107.
Initially, we note that Illinois has been very cautious in its approach to the admission of evidence derived from the use of dogs. (See People v. Pfanschmidt (1914), 262 Ill. 411, 104 N.E. 804; People v. Cruz (1994), 162 Ill. 2d 314, 643 N.E.2d 636.) In Cruz, the Illinois Supreme Court stated that this caution is based both on the fallibility of dogs and the "superstitious awe” with which people view dog-sniff evidence. Cruz, 162 Ill. 2d 314, 643 N.E.2d 636.
The record in this case outlines the differing viewpoints of professionals in the field of arson investigation regarding whether any limitations should be placed on the use of accelerant-sniffing dogs. One faction, led by the IAAI, believes that evidence of a dog’s alert should be used only when the dog’s finding has been corroborated by laboratory analysis. This group is led by chemists and other professionals who are not dog handlers. Others, composed mainly of dog handlers and led by the CADA, contend that an alert *1034may be used without laboratory confirmation as another piece of the evidentiary puzzle.
This disagreement leads us to conclude that the trial court was correct in barring the State’s evidence. There is no "general acceptance” of the reliability of uncorroborated alerts in the field of arson investigation. While dog handlers believe that canine alerts are more reliable than laboratory tests, chemists and other investigators are more skeptical of the dogs’ abilities. As the CADA points out, this conflict may be no more than each faction’s attempt to preserve its own sphere of influence; yet, without general acceptance in the arson investigation field, dog alerts that are not confirmed by laboratory analysis do not meet the Frye test and are properly barred.
For the foregoing reasons, the judgment of the circuit court of Rock Island County is affirmed.
Affirmed.
McCUSKEY and SLATER, JJ„ concur.