Abstract.-I used comparative and experimental analysis of egg size in a Sceloporus lizard to examine a fundamental tenet of life-history theory: the presumed trade-offs among offspring number, offspring size, and performance traits related to offspring size that are likely to influence fitness. I analyzed latitudinal and elevational patterns ofegg life-history characteristics among populations and experimentally manipulated egg size and hatchling size by removing yolk from the eggs to examine the causal bases of population differences in offspring traits.Mean clutch size among populations increased to the north (seven vs. 12 eggs/clutch, California vs. Washington), whereas eggsize decreased (0.65 g vs. 0.40 g). The elevational patterns in southern California paralleled the latitudinal trends. Several offspring life-history traits that are correlated with egg size also varied geographically; these traits included incubation time, hatchling size, growth rate, and hatchling sprint performance.Hatchling viability of experimentally reduced eggs was remarkably high (-70%), even when up to 50% of the yolk was removed. The experimentally reduced eggs and hatchlings demonstrated the degree to which size influences each of the offspring life-history traits considered. Northern eggshatched sooner, in part because oftheir small size. Though growth rate is allometrically related to size within each population (i.e., smaller hatchlings grow faster on a mass-specific basis), population differences in growth rate, as measured in the laboratory, are likely to reflect genetic differentiation in the underlying physiology of growth. Moreover, smaller juveniles, because of experimental reduction, had slower sprint speeds than larger juveniles. The slower sprint speed of hatchlings from Washington compared to hatchlings from California is thus largely due to the fact that eggs are smaller in the Washington population.These results provide a basis for interpreting the evolutionary divergence of the suite of traits involved in the evolution ofmaternal investment per offspring in lizards. For example, evolutionary divergence in some offspringtraits functionally related to size (e.g., sprint speed) may be constrained, relative to traits that are determined by other aspects of development or physiology (e.g., growth). I also discuss issues relating to the evolution of maternal investment that could be tested in laboratory and natural populations using experimentally reduced offspring.