Breeding records from a laboratory colony of Peromyscus maniculatus bairdii were used to examine patterns of covariation among variables describing reproduction, size, age, parity, and survival. Litter size was strongly negatively correlated with average weights of young at birth and at weaning, but positively related to the total mass of young. Mother's weight was correlated with the number of young in a litter and the average and total weights of young at birth and weaning, though the strength of the correlation varied according to when young were weighed. Weight gained by young during the nursing period also was positively related to mother's size. Parity strongly affected litter sizes and birth weights; however, this effect was attributable to continuing growth of the mother. Similarly much, but not all, of the effect of parity on weaning weights and growth rates was attributable to mother's weight. Female age (independent of weight and parity) was important in determining the number of young born (but not their birth weight), the rate of growth of young during the nursing period, and the average weights of individual young at laboratory weaning. The size of young at birth and their number of littermates was clearly related to their probability of surviving the nursing period, and size differences among young at weaning persisted well into adult life. Parity had no effect on the survival of young. The duration of gestation also was related to the weight of a mother and to the weight of her previous litter; the size of the litter in utero had no significant effect on gestation length. These relationships are discussed from the points of view of (1) the advantage to females and their young of large size, (2) the compromises a female makes in determining the number, size, and timing of birth of her offspring, and (3) the implications of these patterns of covariation for investigations of the evolution of reproductive patterns.