Noncomminuted hair samplings (100 mg) from 132 individuals were analyzed for cocaine products by both RIA and GC/MS with D3 internal standards and selected ions for cocaine, benzoylecgonine, and methylecgonine. Ethanol and pH 7 buffer washing until the washes were negative by RIA or GC/MS were followed by overnight digestion in warm 0.1M HCI. A small portion of digest buffered to pH 7 was analyzed by RIA. The remainder, internally standardized, buffered extract was TMS derivatized and analyzed on a capillary OV-1 column by GC/MS. Of the 132 specimens, 10 were positive by RIA and positive by GC/MS for cocaine only, while another 20 specimens were positive for both cocaine and benzoylecgonine by GC/MS as well as positive by RIA. One specimen that was positive by RIA was positive for cocaine and methylecgonine by GC/MS; an additional one was positive for cocaine, methylecgonine, and benzoylecgonine and positive by RIA. The RIA used (100× more reactive to cocaine than to benzoylecgonine) gave positive test results for an additional 47 of the 132 hair samples. Cutoffs used for the RIA were equivalent to 2 µg benzoylecgonine or 0.04 µg cocaine/g hair, and for GC/MS, cutoffs per gram hair were 0.1 µg cocaine and 0.2 µg benzoylecgonine or methylecgonine. Relationships between RIA and GC/MS results and distribution of values found are discussed, as are factors that appear to affect recovery from the hair.