At the appellant's trial for aggravated assault, an expert witness testified that cartridge cases from unfired bullets found in the appellant's apartment had distinct marks that matched fired cartridge cases found at the scene of the offense. We granted review to determine whether the expert's testimony met the reliability requirement in Kelly v. State, 824 S.W.2d 568 (Tex. Crim. App. 1992). We conclude that it did not and reverse the Court of Appeals's judgment.

On direct appeal, the appellant complained that the trial court abused its discretion when it allowed a firearms and toolmark expert to testify concerning magazine marks left on cartridge cases when the State did not show that the scientific testimony was sufficiently reliable under Kelly v. State, 824 S.W.2d 568 (Tex. Crim. App. 1992). The Court of Appeals disagreed. It held that the State had met its burden of showing the reliability of the expert's theory and technique and that the trial court did not abuse its discretion in allowing the testimony. Sexton v. State, 12 S.W.3d 517, 521 (Tex. App.-San Antonio 1999).

We granted review to determine whether the State met its burden of showing the reliability of Crumley's testimony. ⁽¹⁾ We will begin with the facts of the case.

On August 24, 1997, three people were shot while waiting at a stoplight in San Antonio. They could not identify the person who shot them, but were able to describe the assailant's vehicle. The police recovered sixteen fired cartridge cases from the scene, four of which were nine millimeter cartridge cases.

The investigation led the police to Chris Aguero, who owned a vehicle that matched the description of the assailant's vehicle. Aguero confessed to being the driver in the assault, and, in exchange for a plea bargain, identified the appellant as one of the shooters. The police went to the appellant's home, where they received permission to search from the appellant's father. During the search, the police recovered twenty-four live nine millimeter bullets from the appellant's bedroom.

The twenty-four live bullets and the cartridge cases recovered from the scene of the assault were given to Ronald Crumley, a Bexar County firearms and toolmark expert, for analysis. Crumley identified two sets of marks on the live rounds and the fired cartridge cases. According to Crumley, there were twelve live rounds that had distinct marks that matched two of the fired cartridge cases. Another group of twelve live rounds had distinct marks that matched two other fired cartridge cases. Based on these marks, Crumley concluded that the marks found on the live rounds and the cartridge cases were made by a magazine or magazines, and that the fired cartridge cases and the live cartridge cases had at one time been in the same magazine or magazines. ⁽²⁾ The police arrested the appellant and charged him with aggravated assault with a deadly weapon.

Before trial, the appellant filed a motion objecting to Crumley's scientific testimony that the fired cartridge cases found at the scene of the crime and the unfired shells found in the appellant's home had been cycled through the same magazine or magazines. At the hearing on the motion, Crumley was the only witness to testify. He explained that he based his conclusion on firearm and toolmark theories. The general theory behind toolmark examination is that harder metals will leave marks on softer metals when they come in contact with each other. The lips of a magazine, the part of a magazine that keeps the bullets in place, may come into contact with the softer shell casing and leave a mark.

Crumley testified that the marks left on a cartridge from a magazine are unique to that magazine, like a fingerprint. If the magazine marks left on the cartridge cases are sufficiently clear, a firearms and toolmark expert may determine, with the use of a comparison microscope, whether the shells had been cycled through the same magazine. Crumley testified that, if two cartridge cases share the same magazine mark, then one could say with one hundred percent certainty that the two cartridge cases had been cycled through the same magazine. Based on his training and analysis with the comparison microscope, Crumley concluded that the cartridge cases recovered from the crime scene and the unfired cartridge cases found in the appellant's home had been cycled through the same magazine or magazines.

To support his conclusions, Crumley testified that his supervisor, Ed Love, had examined the same live shells and cartridge cases and had reached the same conclusion. Furthermore, Crumley cited treatises that mention that magazines may make individual marks on a cartridge case. Crumley also testified that although he had identified such marks during his training at the Department of Public Safety, this was the first time he had been asked to make an identification with the use of only magazine marks.

The trial court denied the motion and found Crumley's testimony reliable. Crumley testified at trial, and the jury found Sexton guilty of aggravated assault.

We review a trial court's decision to admit or exclude scientific expert testimony under an abuse of discretion standard. See Weatherred, 15 S.W.3d 540, 542 (Tex. Crim. App. 2000). If the trial court's ruling is within the zone of reasonable disagreement, then the trial court's ruling will be upheld. See id.; Kelly, 824 S.W.2d at 579.

Rule of Evidence 702 dictates, "If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise." Tex. R. Evid. 702.

Under Rule of Evidence 702, the trial court serves as a gatekeeper and determines whether the proffered scientific evidence is sufficiently reliable and relevant to aid the jury. ⁽³⁾ Jackson v. State, 17 S.W.3d 664, 670 (Tex. Crim. App. 2000); Kelly, 824 S.W.2d at 573. The proponent of the scientific evidence must demonstrate through clear and convincing evidence that the evidence is in fact reliable. Jackson, 17 S.W.3d at 670; Weatherred, 15 S.W.3d at 542; Kelly, 824 S.W.2d at 573. The proponent of the evidence meets this burden by showing that: (1) the underlying scientific theory is valid, (2) the technique applying the theory is valid, and (3) the technique was properly applied on the occasion in question. Kelly, 824 S.W.2d at 573.

To aid its determination of reliability, the trial court may refer to seven non-exclusive factors: (1) the extent to which the underlying scientific theory and technique are accepted as valid by the relevant scientific community, if such community can be ascertained, (2) the existence of literature supporting or rejecting the underlying scientific theory and technique, (3) the clarity with which the underlying scientific theory and technique can be explained to the court, (4) the potential rate of error of the technique, (5) the availability of other experts to test and evaluate the technique, (6) the qualifications of the expert(s) testifying, and (7) the experience and skill of the person(s) who applied the technique on the occasion in question. Ibid.

Because the appellant challenges the determination that Crumley's testimony was reliable, we will review the evidence produced at trial according to the Kelly criteria and factors. The underlying scientific theory is toolmark theory. The technique in this case was the use of magazine marks to connect items found at the scene of the offense with items found in the appellant's apartment.

We will combine the first two Kelly factors, acceptance by the relevant scientific community and the existence of literature supporting or rejecting the theory and technique. The Court of Appeals noted that toolmark theory for the identification of fired bullets enjoys widespread acceptance and that Crumley, in his testimony, mentioned three sources discussing the identification of fired cartridges from toolmarks. Sexton, 12 S.W.3d at 520.

Two of the sources mentioned by Crumley ⁽⁴⁾ make passing reference to magazine marks, but these are included in discussions for identifying bullets that have been fired through a weapon where there are several marks for comparison made by several components of the firearm. See Vincent Dimaio, Gunshot Wounds: Practical Aspects of Firearms, Ballistics and Forensic Techniques 31 (1985); Julian S. Hatcher, Firearms Investigation, Identification and Evidence flyleaf, 310 (1957) (revision of Hatcher's Textbook of Firearms Investigation, Identification and Evidence, published in 1935). These references may support the general theory of using toolmarks for connecting cartridges generally, but it does not support the particular application in this case.

The first two factors, acceptance by the relevant scientific community and the existence of literature supporting or rejecting the theory and technique, weigh against finding the testimony reliable.

The clarity with which the underlying theory and technique were explained to the trial court weighs in favor of admission. Crumley, by way of analogy, explained that this technique was similar to identifying fingerprints. This is a concept that is easily understandable by laypeople.

The potential rate of error factor weighs in favor of exclusion. Crumley testified that this technique is one hundred percent accurate. Vincent DiMaio, author of one of the treatises on which Crumley relied, testified that he could not be one hundred percent accurate in matching two cartridge cases based on their magazine markings unless he had the actual magazine. The Court of Appeals found it significant that DiMaio's interest in firearms identification was only a hobby. DiMaio, the Court of Appeals noted, did not qualify as a firearm and toolmark expert, ⁽⁵⁾ therefore his opinion should not discount the reliability of Crumley's assertion that the technique employed was one hundred percent accurate.

Crumley's bare assertion is one that the available literature contradicts, however. The literature says that these marks may enable an examiner to connect cartridge cases with the same weapon. The only literature that explains what circumstances make it possible for an examiner to do so requires that the examiner possess knowledge of the manufacturing process of the tool surface and have the tool available for creating test toolmarks. In this case, the magazine or magazines that made the marks upon which Crumley based his identification were not found by the police. Therefore Crumley was not able to make test marks for comparison. Also, Crumley did not say whether he was familiar with the manufacturing process of the magazine or magazines that he said left identifiable marks on the live rounds and cartridge cases.

The record contains evidence of Crumley's qualifications in the area of firearms identification generally. The Court of Appeals summarized Crumley's testimony as follows.

Crumley testified he has worked as a firearm and toolmark examiner for over five years. He trained at the Texas Department of Public Safety (DPS) crime lab in Austin for one-and-a-half years. As a part of his DPS training, Crumley learned to identify magazine marks and match cartridges based on these marks. He also attended training programs by the Southwest Institute of Forensic Science and the Association of Firearm and Toolmark Examiners, and he taught courses on toolmark examination at the DPS crime lab, the Houston Police Department, and the University of Texas Police Department. Crumley has also written three articles on firearm and toolmark examination, each published in the Association of Firearm and Toolmark Examiners' Journal. In total, he has testified as a firearm and toolmark expert in almost fifty cases. However, this was the first case in which he examined magazine marks and testified about the results.

Sexton, 12 S.W.3d at 520. This record qualifies Crumley as a firearms identification expert, but does not support his capacity to identify cartridge cases on the basis of magazine marks only.

We conclude, based on the record before us, that the underlying theory of toolmark examination could be reliable in a given case, but that the State failed to produce evidence of the reliability of the technique used in this case. Under the Kelly criteria, the State failed to show that the technique applied in this case was valid.

The judgment of the Court of Appeals is reversed, and the case is remanded to that Court to perform a harm analysis under Tex. R. App. P. 44.2 (b).

Delivered: October 9, 2002.

Publish.

1. The specific ground on which we granted review was: "The Court of Appeals, in affirming the trial court's decision to admit expert testimony involving magazine marking identification of unfired [cartridge cases] without any other supporting toolmark examination, sets a precedent for the admission of unreliable expert testimony which should be excluded under the holdings of Daubert and Kelly."

2. Crumley explained that he could not tell for sure whether all of the live shells and cartridge cases had been in one magazine together or whether half had been in one magazine and half had been in a second magazine. He testified that there are some magazines, because of their design, that would leave one mark on half of the shells, and a separate mark on the other half.

3. Because the appellant does not challenge the relevancy of the evidence, we will address reliability only.

4. We were unable to obtain a copy of the AFTE [Association of Firearm and Toolmark Examiners] Training Manual that Crumley mentioned in his testimony, and the State did not provide a copy of the relevant reference.

5. As a result, we find its reliance on the treatise written by DiMaio to be unusual.