Invasive species are a major threat and primary driver of vertebrate extinctions on islands, including native and endemic lizards such as West Indian Groundlizards (Genus Pholidoscelis). The genus comprises 19 extant species that range collectively from the Bahamas through the Greater Antilles south to Dominica in the Lesser Antilles. Few studies have addressed aspects of Groundlizard population ecology despite invasive mammals being implicated in Groundlizard declines and three extinctions. From 2007 to 2011, we conducted a capture–recapture study of Puerto Rican Giant Groundlizards (Pholidoscelis exsul) on Guana Island. Guana Island is privately owned and hosts a small number of feral cats, but is mongoose-free and has otherwise experienced only minor human modification since the mid-18th Century. Thus, the island provides a unique opportunity to study lizard demography under conditions minimally impacted by human activities. Our objective was to establish benchmark demographic parameter estimates of P. exsul that may be used in future studies to detect ecological changes. We hypothesized that lizards would be encountered more frequently in disturbed habitat than in less-disturbed habitat because of available anthropogenic food subsidies. We also predicted that large lizards would have higher annual survival rates than would small lizards, and lizards with higher indices of body condition would have correspondingly higher annual survival rates than lizards of the same size class with lower indices of body condition. Our results showed that male growth rates were comparable to similar studies, and individuals reached the presumed size of sexual maturity within 1 yr, with growth rates declining thereafter. Annual survival increased with size class (small = 0.155, 95% confidence interval [CI] = 0.069–0.312; large = 0.518, 95% CI = 0.374–0.660), but we found no evidence of a body condition effect on annual survival. The probability of mortality increased to 99.4% by age 6, but evidence of senescence via declining body condition was apparent sooner. Daily capture probabilities were higher in heavily modified habitat (0.081, 95% CI = 0.049–0.132) than in relatively natural habitat (0.025, 95% CI = 0.012–0.054). Mean maximum distance moved between recaptures was 24.4 m (2–350 m). Estimates of annual abundance ranged from 139 to 250 individuals. Density estimates among years based on the upper confidence limit ranged from 41.9–81.5 individuals/ha within the core sampling area (6.01 ha) and 20.1–39.2 individuals/ha when a 12.5-m buffer strip was incorporated into the effective sampling area.