The number of nestlings in tree-nesting raptors is usually determined during ringing However, an unknown proportion of nests is watched from the ground only. This methodological study compares the number of nestlings as determined by ground observations with the actual number of nestlings (as recorded during ringing). In all species concerned, discrepancies between both types of observations were found, especially in Goshawk (in 20 out of 57 nests the number of nestlings was incorrectly estimated by ground observations) and Sparrowhawk (29 out of 38). Both species are highly sexually dimorphic, with sometimes pronounced agedifferences within a single nest and sex-related differences in growth strategy (males have a faster plumage and behavioural development, thus fledge at an earlier age). Much smaller discrepancies between ground observations and ringing controls were found in Honey Buzzard (8 out of 58 incorrect) and Common Buzzard (7 out of 89 incorrect) Of Kestrel (3 out 5 incorrect) and Hobby (1 out of 6 incorrect), the number of nests was too small to draw conclusions. Ground observations tended to become more unreliable with increasing nesting height, especially in nests in Douglas fir and Giant fir (Appendix 1). Moreover, when many nestlings were present, as in Sparrowhawk, discrepancies of up to three nestlings were found (Table 2). Also in Goshawk (Table 1) and Common Buzzard (Table 3) ground observations tended to become more unreliable with increasing number of nestlings. In conclusion, estimating the number of nestlings from the ground is unreliable and if a nest control is not otherwise possible, it should be specifically stated on the nest card that the number of nestlings is based on ground observations