Principal component analysis and cluster analysis were applied to 2 data sets-65 mammal species (14 orders, 30 families, 10 traits) and 162 mammal species (22 orders, 64 families, 6 traits). In both data sets, most of the covariation in life-history traits could be accounted for by the first 2 factors, and in both data sets their interpretation was similar. The 1st factor, accounting for 68-75% of covariation of life-history traits, ranked the animals from early maturing, small, short-lived animals with short gestation periods, many small altricial young, and long lactation periods, to the opposite suite of traits in large animals. The 2nd factor, accounting for 12-20% of the covariation in life-history traits, ranked the animals from altricial to precocial. When order and family effects were removed, the proportion of variance accounted for by the 1st factor dropped 29-36%, and the number of high level clusters increased. Weight alone accounts for much of the tendency for mammals to be arranged on a single axis ranging from early maturing, highly fecund, and short-lived small animals to the opposite, but after the effects of weight were removed, order and family effects still had significant impact on patterns of covariation. Thus, morphological design, in addition to size, may constrain life-history evolution.-from Author