Longitudinal studies of senescence accumulate rapidly from natural populations. However, it is largely unknown whether different fitness components senesce in parallel, how reproductive and survival senescence contribute to declines in reproductive value, and how large the fitness cost of senescence is (the difference between the observed reproductive value and the hypothetical reproductive value, if senescence would not occur). We analyzed age-specific survival in great tits Parus major and combined our results with analyses of reproductive senescence to address these issues. Recapture probability of breeding females declined with age, suggesting age-specific increases in skipped or failed breeding and highlighting an important bias that studies of senescence in wild populations should incorporate. Survival probability also declined with age and in parallel with recruit production. Reproductive value decreased 87% between age 1 and age 9 but at a fitness cost of only 4%; the proportion of the contribution of reproductive senescence versus survival senescence to this cost was 0.7. For 11 other species, we estimated fitness costs of senescence of 6%-63% (average: birds, 9%; mammals, 42%), with relative contributions of reproductive senescence of 0.0-0.7 (average: birds, 0.4; mammals, 0.3). We suggest that understanding when and why reproductive and survival senescence differ will help in the identification of proximate mechanisms underlying variation in rates of senescence and its evolution.