North American river otters (Lontra canadensis) were trapped in a managed marsh in northern California between July and November of 1998. Five trap types using several set configurations were utilized in this study. Otters were successfully captured with minimal injury and a relatively high trap rate (1 capture per 48 trap nights), with the use of No. 1.5 and No. 1.75 double-coil spring traps, and No. 11 double long spring traps on short chains in blind land sets, or in bank sets on one-way cables leading to land-buried stakes. Only 3 of 14 captures incurred more than minimal injuries, and all 3 of these were exacerbated by complications from traps attached to long chains. Otters were captured at any time of day, although activity appeared lowered between 0900 and 1900 hours. Otters were anesthetized for handling. Monitoring results and anesthetic complications are reported for 14 captures. Ketamine (15 mg/kg) and midazolam (0.5 mg/kg) provided a wide margin of safety, rapid induction at low injection volume, good anesthetic quality, handling time of about 30 min, and few complications. White blood cell counts were taken and appeared high relative to reported values in other studies. This trend is likely attributable to blood draws immediately after the animal had been held in a trap for several hours, in contrast to most other studies, in which blood was drawn days to weeks after being held in captivity.