The question of assessing renal toxicity in studies of humans exposed to environmental toxicants is of major concern in health effects studies. The measurement of urinary enzymes is a reliable, sensitive indicator of renal tubular damage due to a variety of causes, including heavy metal and drug toxicity. However, urine is not a favorable matrix for some enzymatic activities because certain enzymes are less stabile in urine than in a matrix such as serum. This study addresses the problem of the lack of stability of alanine aminopeptidase (AAP) and gamma glutamyltranspeptidase (GGT) in the urinary matrix. Four urinary treatment methods consisting of dialysis or the addition of albumin, dimethyl sulfoxide, or glycerol have been examined and compared. The activities of AAP, GGT, andN-acetyl-Β-D-glucosaminidase (NAGA) in six normal treated and untreated urines were followed for 28 days at 4‡C and for 6 months at −20‡C. This study indicates that of the methods tested, giycerol treatment provides the most stable AAP and GGT activity and is the most reliable method for stabilizing these enzymes in frozen urine. Even with giycerol treatment, some GGT activity is lost (mean loss of 11% over 6 months at −20‡C), indicating GGT should be assayed as soon as possible. NAGA activity is most stable in untreated urines but still increases by approximately 9% (mean of six samples) in 6 months at −20‡C. The implications of observed effects of dialyzable inhibitors are discussed for the three enzymes.