We reviewed information on the demands of incubation to examine whether these could influence the optimal clutch size of birds. The results indicate that appreciable metabolic costs of incubation commonly exist, and that the incubation of enlarged clutches can impose penalties on birds. In 23 studies on 19 species, incubation metabolic rate (IMR) was not elevated above the metabolic rate of resting non-incubating birds (RMR), but contrary to the physiological predictions of King and others, IMR was greater than RMR in 15 studies on 15 species. Across species, IMR was substantially above basal metabolic rate (BMR), averaging 1.606×BMR. Of six studies on three species performed under thermo-neutral conditions, none found IMR to be in excess of RMR. IMRs measured exclusively within the thermo-neutral zone averaged only 1.08×BMR contrasting with the significantly higher figure of 1.72×BMR under wider conditions. 16 of 17 studies on procellariiforms found IMR below RMR, indicating a significant difference between this and other orders. We could find no other taxonomic, or ecological factors which had clear effects on IMR. Where clutch size was adjusted experimentally during incubation, larger clutches were associated with: significantly lower percentage hatching success in 11 of 19 studies; longer incubation periods in eight of ten studies; greater loss of adult body condition in two of five studies; and higher adult energy expenditure in eight of nine studies. Given that incubation does involve metabolic costs and given that the demands of incubation increase sufficiently with clutch size to affect breeding performance, we propose that the optimal clutch size of birds may in part by shaped by the number of eggs the parents can afford to incubate.