Excretion rate estimates are important for linking consumers to biogeochemical processes. Short-term incubations in chambers are a common approach for studies. This, however, may result in inaccuracies due to a welldocumented decline in excretion with time, which is often attributed to fasting. An alternative explanation, however, is that excretion slows during recovery from handling stress. Whereas shorter incubations may reduce fasting effects, longer incubations can allow for recovery from handling. Although both views have merit and are not mutually exclusive, attempts to identify optimal incubation times are lacking. We examined effects of handling stress and fasting on NH 4 + excretion by omnivorous tadpoles and predatory fish. Fasting did not influence tadpole excretion for at least 6 h. Rather, handling stress appeared to cause initially elevated excretion, and thus longer incubations were appropriate. Fasting and handling stress influenced fish excretion, and intermediate incubations (~30 min) appeared most appropriate. We also found that ammonium measurements were confounded by increasing concentrations of excretory products in longer incubations. We recommend that studies include incubation time trials and correct for potential matrix effects. Effects of handling and fasting likely vary with organism and experimental procedures, and both can be accounted for using the procedures we describe.