From 1985 through 1990, I monitored the populations of two species of toads, Bufo quercicus and B. terrestris, at a temporary pond in the xeric uplands of north-central Florida. A drift fence with pitfall traps completely encircled the pond basin; the fence was monitored 5 days per week throughout the year. The 5-year study coincided with a severe regional drought that resulted in generally short hydroperiods at unpredictable times of the year. More than 800 toads were captured. Successful metamorphosis never occurred at the pond although toads continued to visit it throughout the study. The sex ratio was male biased in B. quercicus but not in B. terrestris, although significant variation was observed from one year to the next. Likewise, the size-class structure and length-weight patterns varied among species, sexes, and years. Although fewer toads entered the pond basin as the study progressed, toads may have gone elsewhere to breed or they may have remained in refugia. Thus, decreased capture does not necessarily indicate that a drought-related population decline occurred. Drought may have disrupted normal arrival patterns and length of stay within the pond basin. Drought also could be responsible for variation in annual size-class structure of captured toads. The uncertainty of the hydroperiod both spatially and temporally in adjacent breeding sites, the ability of toads to move long distances with the potential for migration between breeding sites, and the lack of specificity in the choice of breeding sites (i.e. permanent versus different types of temporary wetlands) may lead to the formation of metapopulations in the xeric upland habitats of north-central Florida. Long-term monitoring under a variety of climatic conditions is needed to assess the effects of drought and other types of environmental stresses on toad populations.