Global change, including habitat isolation and climate change, has both short‐ and long‐term impacts on wildlife populations. For example, genetic drift and inbreeding result in genetic impoverishment in small, isolated populations, while species undergo range shifts or adaptive phenotypic change in response to shifts in environmental temperatures. In this study, we utilize a model system in which Holocene landscape changes have occurred to examine long‐term effects of population isolation. To examine how isolation may constrain responses to climate change, we characterized ecophysiology across land‐bridge island populations of Erhard's wall lizard Podarcis erhardii. We hypothesized that 1) small, isolated populations that are likely genetically depauperate would exhibit lower phenotypic variability; and 2) populations would be adapted to local microhabitat conditions. We compared a population at a low elevation site on the large island of Naxos with two small populations on nearby islets to determine the effects of population fragmentation. We further compared the low elevation Naxos population with two high elevation sites characterized by disparate microclimates to examine the effects of microclimate. To assess the thermal biology and ecophysiological limits of the study species we measured operative body temperatures (Te), field body temperatures (Tb), preferred temperatures (Tp), thermal tolerances (CTmax and CTmin), and evaporative water loss (EWL). Our results indicate that small, isolated populations did not exhibit thermal biology or evaporative water loss, while EWL and thermoregulatory effort varied according to microhabitat characteristics. This study integrates fine‐scale measurements with environmental data to provide a holistic view of the relationships between ecophysiology, fragmentation, and microclimate. Our methods can be applied to other ectotherms to gain a better understanding of potential impacts of global change on natural populations.