We test the hypotheses proposed by Gentry (1991) and Schnitzer (2005) that liana density and basal area in tropical forests vary negatively with mean annual precipitation and positively with seasonality. Previous studies correlating liana abundance with these climatic variables have produced conflicting results warranting a new analysis of drivers of liana abundance based on a different dataset. We compiled a pan-tropical dataset containing 28953 lianas (≥ 2.5 cm diameter) from studies conducted at 13 Neotropical and 11 Paleotropical dry to wet lowland tropical forests. The ranges in mean annual precipitation and dry season length (number of months with mean rainfall < 100 mm) represented by these datasets were 861–7250 mm/yr and 0–7 months respectively. Pan-tropically liana density and basal area decreased significantly with increasing annual rainfall and increased with increasing dry season length supporting the hypotheses of Gentry (1991) and Schnitzer (2005). Higher understory light regimes and traits such as deep root systems and high water-use efficiency may enable lianas to proliferate in drier tropical forests. Our results suggest that much of the variation in liana density and basal area in the tropics can be accounted for by using the relatively simple metrics of mean annual rainfall or dry season length.