The polymerase chain reaction (PCR) was used to amplify the HLA DQ alpha gene using DNA recovered from evidentiary samples. Amplified HLA DQ alpha DNA was then typed using sequence-specific oligonucleotide probes. Slight modifications of previously published DNA extraction methods improved typing success of bloodstains and semen-containing material. Evidentiary samples, consisting of 206 known bloodstains, 26 questioned bloodstains, and 123 questioned semen-containing evidentiary materials were analyzed from 96 cases previously analyzed by restriction fragment length polymorphism (RFLP) typing in the FBI Laboratory. Of the known bloodstains, 98.5% yielded DQ alpha typing results. Of the questioned samples, 102 of 149 (24/26 bloodstains and 78/123 semen-containing materials), or 68%, produced typing results. Of the 78 cases that were RFLP inclusions, 59 yielded interpretable DQ alpha results and these were all inclusions. The remaining 19 cases could not be interpreted for DQ alpha. Of the 18 RFLP exclusions, eleven were DQ alpha exclusions, four were DQ alpha inclusions, and three could not be interpreted for DQ alpha. It is expected that because of the difference in discrimination potential of the two methods, some RFLP exclusions would be DQ alpha inclusions. Some samples that failed to produce typing results may have had insufficient DNA for analysis. Employment of a human DNA quantification method in DQ alpha casework would allow the user to more consistently use sufficient quantities of DNA for amplification. It also could provide a guide for determining if an inhibitor of PCR is present, thus suggesting the use of a procedure to improve amplification. This study provides support that the HLA DQ alpha typing procedure is valid for typing forensic samples.