Although growing urbanization has direct negative consequences for local biodiversity, several native species have been observed maintaining populations in urban environments. Understanding which factors influence the ability of native species to persist in urban environments is crucial, both for the study of biological adaptation and of urban planning. The quantification of the proportion of juvenile individuals can be a good proxy for assessing the long-term persistence of urban populations. We present comparative data about spatial and temporal variations in the age-class structure in two suburban and two forest populations of the Cuban endemic lizard Anolis homolechis, obtained during a 20-month survey. We found a four-fold lower proportion of juveniles in the suburban habitat compared to the forest one. There was, however, no evidence for differential female fecundity between the two habitats, as assessed by the proportion of gravid females. Conversely, the rate of tail autotomy (an antipredator behavior) was significantly higher in the suburban juveniles compared to the forest ones, possibly reflecting a higher exposure to predators and, particularly, inter- and intraspecific cannibalism. However, tail loss at initial capture or habitat type had no effect on the probability of recapture of juveniles. We discuss the potential causes and consequences of a modified age-class structure in urbanized environments.