The Netherlands holds internationally important numbers of Black-tailed Godwits Limosa limosa, and reports of a substantial population decline have prompted concern. One way of narrowing the list of possible causes of this decline is to identify the demographic processes responsible. For this reason, we conducted a survival analysis for the period 1960 to 2000 based on dead recoveries of godwits. Although declining sample sizes and falling reporting rates reduced the precision of estimates in later years, we concluded that adult survival had dropped significantly in the late 1990s from 0.77 to 0.40. Data were too sparse to estimate juvenile survival in this period. Apart from this sharp drop, we found no evidence of long-term trends in either first-year or adult survival. We found no evidence for effects on survival estimates of shifts in the spatial distribution of recoveries or in the reported causes of death. These large-scale analyses of birds recovered dead complement more intensive mark-resighting studies at individual sites. Given the falling reporting rates of dead birds and consequent loss of precision of survival estimates, intensive mark-resighting programmes have an increasingly important role to play in the study of survival rates.