We radio-marked 150 male American woodcock (Scolopax minor) during 1987–1989 and estimated period survival for 1 April – 15 June. Survival varied from 0.690 (1989) to 0.924 (1988), with a 3-year mean (95% confidence interval) of 0.789 (0.693–0.885). Woodcock were killed by raptors (n = 14, 53.8%), mammals (n = 1, 3.8%), or unknown predators (n = 5, 19.2%); six deaths (23.1%) were from miscellaneous causes, including three (11.5%) from entanglement in the transmitter harness. A composite survival estimate based on telemetry studies for the breeding, postbreeding, and wintering periods was 0.471 (0.789 × 0.923 × 0.647). The calculated survival rates were 0.881 for the spring migration period and 0.853 for the combined hunting and fall migration period. In a proportional hazards model, body mass at capture was not related to survival. Forest type (hardwood versus conifers) affected survival (P < 0.016), which was lower for woodcock using mostly conifer sites. Survival was related positively to mean snow depth in December (P < 0.038), negatively to snow depth in April (P < 0.046), and positively to minimum temperature in December (P < 0.054) and April (P < 0.066) in some analyses.