The reproducibility of the micromodification of the Mills procedure for detection of six pesticide residues in biopsied adipose tissue was investigated. The magnitude of the pooled standard deviations implies that the measurements should be rounded to the nearest tenth of a ppm for DDT, DDE, and benzene hexachloride, and to the nearest hundredth of a ppm for DDD, dieldrin, and heptachlor epoxide. Although variability among measurements of certain pesticides by a tritium detector chromatograph was found to be significantly greater than among measurements by nickel detector chromatograph, no inferences can be made from this sample of only two chromatographs.