Competition between the insectivorous iguanid lizards Sceloporus merriami and Urosaurus ornatus was studied in a density manipulation experiment conducted in the Grapevine Hills of Big Bend National Park, Texas, USA from 1974 through 1977. Resident lizards were individually marked and censused using standard recapture techniques on six experimental plots chosen for similarity in size, habitat, isolation from one another by habitat unsuitable for these lizards, and initial densities of these two species. All S. merriami were removed from two plots, all U. ornatus from two, and two plots functioned as controls. Recaptures of marked individuals provided data on individual growth, age—specific survivorship, size—specific fecundity, and population density. Animals collected near the study plots provided data on size—specific fecundity, individual foraging success, and total prehibernation lipid levels. Observations of marked individuals provided estimates of individual foraging success. Rainfall varied dramatically during the 4 yr of the study; in 1974 and 1976 precipitation was much greater than the 22—yr average, and in 1975 and 1977 precipitation was much less than average. Two methods of estimating arthropod abundance indicated that significantly fewer prey were available to these lizards in the dry years. Individual foraging success, growth rates, body masses, prehibernation lipid levels, and population densities were significantly lower in the control populations of both species in 1975 and 1977 than in 1974 or 1976, indicating that individuals of both species experienced stressful conditions during the dry years. The removal of S. merriami resulted in significant effects on experimental populations of U. ornatus only during 1975 and 1977. The density of U. ornatus on the experimental plots increased significantly after the removal of S. merriami, even though the control populations declined. Individual foraging success, growth rate, prehibernation body mass, and lipid levels were significantly greater in U. ornatus from experimental populations than in those from control populations in the dry years of the study. The two treatments did not differ significantly in the wet years. Experimental populations of S. merriami exhibited significantly higher adult and juvenile survival than did the control populations in 1975. This was the only significant treatment effect demonstrable in experimental populations of S. merriami. Experimental results and data on life history characteristics suggest that these species compete significantly during periods of drought—induced food scarcity but not at other times. The mechanism of this competition is probably exploitative, with the per capita effect of S. merriami on U. ornatus greater than vice versa. Temporal variation in the intensity of competitive interaction between these two species is attributable to variation in food resource abundance, and arthropod abundance in this system was shown to be a function of rainfall. Examination of 22 yr of continuous precipitation data suggests that competition between these species may be frequently absent or undetectable in this system. The results of this study demonstrate that the intensity of interspecific competition can be quite variable in systems such as this and invite caution to the uncritical acceptance of point estimates of the intensity of competitive interaction.