The aims of the present study were (i) to provide basic comparative data on the time course, route, and characteristics of excreted [14C]testosterone (T) metabolites in three nonhuman primates: the common marmoset (Callithrix jacchus), the long-tailed macaque (Macaca fascicularis) and the chimpanzee (Pan troglodytes) and (ii) to use this information to help validate the measurement of urinary and fecal testosterone metabolites for assessing androgen status in Anthropoid primates. Radiolabeled 14C-T (10-30 microCi) was injected intravenously into one adult male of each species and the excreta collected over the next 5 days. Peak radioactivity in urine was detected within 2h and accounted for 67% (Mf), 80% (Cj) and 91% (Pt) of the total radioactivity recovered. The time course of excretion of radioactivity in feces showed a higher variation between species (4-26 h to peak values). In all three species, the majority (>90%) of urinary metabolites were excreted as conjugates whereas the proportion of conjugated metabolites in feces was substantially lower and more variable. High pressure liquid chromatography (HPLC) analysis of urinary and fecal extracts revealed multiple peaks of radioactivity in all three individuals, but each with a distinctive pattern. Native T was excreted in only small amounts into the urine, whereas it was virtually absent in the feces of all three individuals. Three C17 group-specific enzymeimmunoassays using antisera against testosterone, 5alpha-androstane-17alpha-ol-3-one and androsterone were evaluated for their ability to discriminate immunoreactive androgen levels between intact males, castrated males and females based on measurements in urine and feces. In the marmoset, all assays (except for T in feces) clearly discriminated between test groups; in the chimpanzee significantly higher levels of androgen immunoreactivity in intact versus castrated males were measured in urine, but not feces. In the macaque, only the 5alpha-androstanolone measurement in feces discriminated between groups. Data on the results of a radiometabolism study using 3H-DHEA (a weak adrenal androgen) in a long-tailed macaque suggested that co-measurement of metabolites derived from T and DHEA in the assays tested might explain the difficulties in discriminating gonadal status in the two Old World primate species. Collectively, the data show that T metabolism in primates is highly complex and that no single method for noninvasive assessment of androgen status can be used for application across species. The importance of a proper validation of the methodology for each species is emphasised.