Cortisol levels serve as an index of pituitary-adrenal activity in nonhuman primates. In adult monkeys, cortisol is normally measured in blood (typically requiring restraint or sedation) or urine (reflecting a state rather than point estimate). In contrast, saliva collection is less invasive than drawing blood and allows for repeated sampling within a short period of time. Although protocols exist for collecting saliva from young monkeys, these procedures are inadequate for awake, unrestrained adult animals. Our laboratory has developed two methods for collecting saliva from adult rhesus monkeys: a "screen" method, which involves licking screen-covered gauze, and a "pole" method, which involves sucking and chewing on an attached rope. Twenty-three adult male rhesus monkeys were used to evaluate these two methods. After a period of adaptation, saliva samples were collected from 21 of 23 subjects. Saliva collection was faster with the pole than with the screen method (P < 0.01), but the pole method was not suitable for some animals because of their tendency to bite off the attached rope. An analysis of 19 saliva samples revealed a mean cortisol concentration of 0.84 microg/dl (range 0.27-1.77 microg/dl). There was no statistically significant difference in cortisol value between methods used (P > 0.22). The influence of the flavoring on the cortisol assay was tested, and was found to have no significant effect (P > 0.28). Our results indicate that either technique can be used to safely collect saliva from unrestrained adult monkeys. Choice of technique will depend on the proclivities of individual monkeys.