Measures of reproductive success have long been assumed to be reasonable surrogates of fitness. We examined reproductive success and recruitment at the individual and population levels in Swainson's Hawks breeding in northern California. At the population level, we investigated whether the number of individuals subsequently recruited could be predicted by the number of offspring produced in any given year, finding no evidence of differences in probability of recruitment by cohort or year. Instead, age of recruits was the best predictor of probability of their being recruited into the local breeding population. At the individual level, we used generalized linear models to examine the correlation between lifetime reproductive success of individual females and the number of their offspring recruited. For individuals, the number of fledglings produced was significantly correlated with both the number of those offspring and of grandchildren recruited into the breeding population. We also examined the relationship between the average annual reproduction of an individual and the number of offspring recruited, finding these significantly correlated. The relationship was not linear: individuals producing moderate numbers of fledglings, on average, yielded the highest numbers of recruits, reflecting a trade-off between adult survival and reproduction. These results provide evidence that monitoring of reproductive success of individuals or a population can give reasonable indices of future recruitment if the study is continued for a long period. But individuals with the highest average rate of reproduction may not be the most fit.