S., and convictions in many of those cases were based, at least partially, on this particular type of evidence used to connect suspects to bullet fragments associated with the crimes for which they were prosecuted. In some cases, such as that of Lee Wayne Hunt, imprisoned for 21 years for murder in North Carolina, conviction was based exclusively on this type of forensic evidence (Solomon, 2007) on the basis of which prosecutors connected Hunt to ammunition in the possession of a codefendant, establishing Hunt's guilt by virtue of the felony murder rule (Kobalinsky and Liotti, et al., 2005). Comparative Bullet Lead Analysis and Violation of the Scientific Method:

In 1998, William Tobin, a recently retired FBI metallurgical specialist who had previously been credited with identifying the cause of high profile accidents like the 1996 crash of Flight 800 initiated a series of tests in conjunction with a metallurgist at the Lawrence Livermore National Laboratory to test the validity of the underlying assumptions made by FBI chemists using the bullet lead analysis forensic technique. By 2001, that study determined that those initial assumptions were invalid. Specifically, whereas the bureau had postulated that bullet lead chemical composition varied consistently enough to positively identify bullet fragments as having come from a particular box of ammunition, the Livermore study discovered that this assumption was flawed, rendering the forensic technique invalid in many respects.

It turned out that bullets manufactured together from a single batch of lead did not necessarily always match other bullets produced from the same lead batch, by virtue of subtle changes attributable to the manufacturing process (Solomon, 2007). Even more importantly, the study revealed that bullets manufactured more than a year apart sometimes had identical chemical composition with respect to the elements tested by the comparative bullet lead analysis theory.

The researchers had purchased bullets from several different retail stores in one state and discovered that bullets bearing the same manufacturers' production dates distributed within communities as large as an entire state were capable of yielding a match via the...

Those results refuted the scientific validity of the forensic relevance of bullet lead chemical composition analysis at a fundamental level because it meant that the process used to establish guilt in so many previous criminal prosecutions by linking crime scene evidence to ammunition in the possession of defendants was entirely invalid. The bureau has since announced that it would suspend its use of comparative bullet lead analysis at trial and issued letters to hundreds of state and local law enforcement agencies alerting them to the potential invalidity of some of the evidence furnished by the FBI in connection with criminal prosecutions by those agencies (Solomon, 2007).
Conclusion:

In the case of comparative bullet lead analysis, the underlying hypothesis that relative chemical composition of bullets varied measurably on the basis of different batches of lead used in their manufacture - was actually valid, in principle. However, more careful investigation disclosed that (1) the degree of variation was too variable to be relied upon to the degree suggested by FBI testimony, and (2) bullets bearing chemically distinct similarity were incapable of being positively identified as having come from the same box of ammunition, but only as having come from the same general community (Solomon, 2007).

Since the revelation, FBI Director Robert Mueller issued a formal memo in 2005 discouraging state prosecutors from relying on evidence of comparative bullet lead analysis provided by the bureau despite the scientific validity of the method itself.

The problem was with expert testimony that the chemical composition of bullets is as precise as DNA analysis rather than the tests themselves. In the future, comparative bullet lead analysis results may properly be introduced as one of many other factors, but even then, only in the sense that a match indicates a greater likelihood of guilt and not as conclusive evidence of guilt.