In an attempt to describe hormone-behavior interactions in a sample of wild male chimpanzees, we quantified testosterone in 67 fecal samples obtained from 22 adult male chimpanzees at Ngogo, Kibale National Park, Uganda. A mixed-model methodology that controlled for age-class identified a significant positive association between testosterone levels and dominance rank. The results are consistent with those reported from a separate, smaller chimpanzee community in the same population in a study that analyzed testosterone levels in urine [Muller & Wrangham, 2004]. As in that earlier study, our results held during a period of social stability, which is not consistent with predictions of the "challenge hypothesis." We concur with Muller and Wrangham [2004] that the challenge hypothesis requires modification to explain the chimpanzee data, because fission-fusion sociality in chimpanzees makes challenges unpredictable. We also discuss the utility of fecal samples and a mixed-model statistical method for behavioral endocrinology studies.