The growth rate and composition of cave calcite deposits (speleothems) are often used as proxies for past environmental change. There is, however, the potential for bias in the speleothem record due to seasonal fluctuations in calcite growth and dripwater chemistry. It has been proposed that the growth rate of speleothem calcite in Texas caves varies seasonally in response to density-driven fluctuations in cave-air CO 2 , with lower growth rates in the warmer months when cave-air CO 2 is highest. We monitored CO 2 in three undeveloped caves and three tourist caves spread over 130 km in central Texas to determine whether seasonal CO 2 fluctuations are confined to tourist caves, which have been modified from their natural states, and the extent to which cave-air CO 2 is controlled by variations in cave geometry, host rocks, cave volume, and soils. Nearly 150 lateral transects into six caves over three years show that CO 2 concentrations vary seasonally in five of the caves monitored, with peak concentrations in the warmer months and lower concentrations in the cooler months. The caves occur in six stratigraphic units of lower Cretaceous marine platform carbonate rocks and vary in volume (from 100 to .100,000 m 3 ) and geometry. Seasonal CO 2 fluctuations are regional in extent and unlikely due to human activity. Seasonal fluctuations are independent of cave geometry, volume, depth, soil thickness, and the hosting stratigraphic unit. Our findings indicate that seasonal variations in calcite deposition may introduce bias in the speleothem record, and should be considered when reconstructing paleoclimate using speleothem proxies.