To improve understanding of pintail ecology, we radiotagged 191 hatch-year (HY) and 228 after-hatch-year (AHY) female northern pintails (Anas acuta) in the San Joaquin Valley (SJV), and studied their survival throughout central California, USA, during September-March, 1991-1994. We used adjusted Akaike Information Criterion (AICc) values to contrast known-fate models and examine variation in survival rates relative to year, interval, wintering region (AJV, other central California), pintail age, body mass at capture, capture date, capture area, and radio type. The best-fitting model included only interval x year and age x body mass; the next 2 best-fitting models also included wintering region and capture date. Hunting caused 83% of the mortalities we observed, and survival was consistently lower during hunting than nonhunting intervals. Nonhunting and hunting mortality during early winter was highest during the 1991-1992 drought year. Early-winter survival improved during the study along with habitat conditions in the Grassland Ecological Area (EA), where most radiotagged pintails spent early winter. Survival was more closely related to body mass at capture for HY than AHY pintails, even after accounting for the later arrival (based on capture date) of HY pintails, suggesting HY pintails are less adept at improving their condition. Thus, productivity estimates based on harvest age ratios may be biased if relative vulnerability of HY and AHY pintails is assumed to be constant because fall body condition of pintails may vary greatly among years. Cumulative winter survival was 75.6% (95% CI = 68.3% to 81.7%) for AHY and 65.4% (56.7% to 73.1%) for HY female pintails. Daily odds of survival in the cotton-agriculture landscape of the SJV were -21.3% (-40.3% to +3.7%) lower than in the rice-agriculture landscape of the Sacramento Valley (SACV) and other central California areas. Higher hunting mortality may be 1 reason pintails have declined more in SJV than in SACV.