CABL – Compositional Analysis of Bullet Lead – was originally proposed as a means to be able to link a bullet recovered from a crime with a box of ammunition that may be, or may have been, in the possession of a suspect.
The theory was that the precise chemical composition of any particular melt of lead used to make bullets made it unique and distinguishable from all others. The compositional testing was most commonly done by the FBI lab. Many people were convicted of fire-arms-based crimes based upon this theory.
However, when real science started focusing effort on this issue, it was determined that two different melts of lead could, in fact, be compositionally indistinguishable. The major reason for this being that bullets are universally manufactured from melted down car batteries, and the lead used to make car batteries is, by necessity, VERY uniform.
This is another example of where the forensics community presumed “uniqueness” without statistical validation. Once validation was attempted, they were proven wrong.
The FBI has officially acknowledged the discrediting of this evidence, and they ceased doing CABL analysis in 2005. However, there are still cases of wrongful conviction based upon this evidence out there.
Here is some reference material for you:
1) Here is a graphic showing why this is bogus evidence by Laura Stanton, The Washington Post – November 17, 2007
2) In 2004 The (US) National Academies did a study of the validity of CABL. Here is chapter 5 of that report, stating conclusions.
3) Here is a 2007 article by John Solomon of the Washington Post.