A striking feature of reproduction in eutherian mammals is its limited diversity. All have internal fertilization, viviparity, and lactation. Most reproductive traits, including gestation time (Gunther and Guerra, 1955; Kilstrom, 1972; Sacher and Staffeldt, 1976), birth weight (Leutenegger, 1976; Millar, 1977; Robbins and Robbins, 1979), litter weight (Leitch et al., 1959), weight at weaning (Millar, 1977), and post natal growth (Gunther and Guerra, 1955; Millar, 1977; Case, 1978) are highly predictable from adult size. Only offspring number and duration of lactation appear poorly related to adult size (Millar, 1977). A general explanation for these relationships is that they are related to morphological and physiological constraints. Numerous physiological variables, including metabolic rate, are related to adult size (Gunther and Guerra, 1955; Stahl, 1967; Calder, 1974; Gunther, 1975; SchmidtNielsen, 1975). Structural traits are related to mechanical requirements so that they also are predictable from adult size (Huxley, 1932; Bonner, 1965; Gould, 1966). Together, these morphological and physiological patterns appear to influence ecological and behavioral patterns (Sacher, 1959; McNab, 1963; Lindstedt and Calder, 1976; Blueweiss et al., 1978; Harestad and Bunnell, 1979). The existence of a strong design constraint system in mammals presents problems when single traits are examined in relation to environmental conditions; foremost among these problems is that very few traits are independent of other traits. Characteristics not considered in any comparison can lead to biased results. In other words, changes in one trait may occur simply because another trait responds to environmental conditions. Another problem is that very few traits are linearly related to other traits so that the relative magnitudes of traits also vary predictably. Simple ratios cannot be used to compensate for these nonlinear relationships. These and other problems (considered in detail by Gould [1966] and Clutton-Brock and Harvey [1979]) have not always been recognized in studies of mammalian reproductive traits. The purpose of this study was to examine the interrelationships among reproductive traits of mammals in order to define the system of morphological and physiological constraints, and to examine the residual variation in this system in relation to taxonomic groups and ecological habits.