In polytocous mammals an inverse relation is known to exist between the number of young in the litter and the size of the young at birth. It has been customary to attribute this to prenatal competition for a limited pool of nutrients in the maternal circulation (Hammond & Marshall 1952). The present study was undertaken in order to subject the traditional theory to quantitative test. Pregnant mice were killed shortly before term and their foetuses were weighed. Some of the pregnancies had resulted from natural ovulation. Others were obtained by inducing superovulation with hormone treatment, in some cases in adult, and in others in sexually immature females. Analysis of the foetal weights yielded a number of results at variance with the predictions of the theory of competition, notably the following: (1) Foetal growth was more affected by the presence of other foetuses if these were in the same, rather than in the opposite, uterine horn. (2) Foetuses dying in mid-pregnancy, when their nutritional demands would still be small, exerted effects upon the growth of the survivors comparable to those exerted by foetuses surviving the middle period. (3) The signs of nutritional stress exhibited by sexually immature females carrying abnormally large numbers of foetuses, were not accompanied by a reduction in foetal weight after due allowance for foetal number. (4) The foetus occupying the top (ovarian) position in the uterine horn was on average significantly lighter than its neighbour. These results can be better accommodated by an alternative theory of foetal growth (Eckstein, McKeown & Record 1955) which proposes that the chief regulating factor is the pressure at which maternal blood is supplied to the placenta. Some other observations on foetal growth in mice are cited which are also consistent with the haemodynamic theory.