This study presents wing-beat frequency data measured mainly by radar, complemented by video and cinematic recordings, for 153 western Palaearctic and two African species. Data on a further 45 Palaearctic species from other sources are provided in an electronic appendix. For 41 species with passerine-type flight, the duration of flapping and pausing phases is given. The graphical presentations of frequency ranges and wing-beat patterns show within-species variation and allow easy comparison between species, taxonomic groups and types of flight. Wing-beat frequency is described by Pennycuick (J. Exp. Biol. 2001; 204: 3283–3294) as a function of body-mass, wing-span, wing-area, gravity and air density; for birds with passerine-type flight the power-fraction has also to be considered. We tested Pennycuick’s general allometric model and estimated the coefficients based on our data. The general model explained a high proportion of variation in wing-beat frequency and the coefficients differed only slightly from Pennycuick’s original values. Modelling continuous-flapping flyers alone resulted in coefficients not different from those predicted (within 95% intervals). Doing so for passerine-type birds resulted in a model with non-significant contributions of body-mass and wing-span to the model. This was mainly due to the very high correlation between body-mass, wing-span and wing-area, revealing similar relative scaling properties within this flight type. However, wing-beat frequency increased less than expected with respect to power-fraction, indicating that the drop in flight level during the non-flapping phases, compensated by the factor (g/q)0.5 in Pennycuick’s model, is smaller than presumed. This may be due to lift produced by the body during the bounding phase or by only partial folding of the wings.