Tropical ecosystems strongly influence Earth’s climate and weather patterns. Most tropical ecosystems remain warm year-round; nonetheless, their plants undergo seasonal cycles of carbon and water exchange. Previous research has shown the importance of water and light as drivers of the seasonality of photosynthetic activity in the tropics. Although data are scarce, field-based studies have found that seasonal cycles at a handful of tropical forest sites do not match those in land surface model simulations. A comprehensive understanding and model comparison of how seasonal variations in tropical photosynthetic activity relate to climate is lacking. In this study, we identify the seasonal relationships of precipitation and light availability with satellite-based photosynthetic activity. Three dominant and spatially distinct seasonal relationships emerge between photosynthetic activity and these two environmental drivers: photosynthetic activity that is positively correlated with both drivers (36% of tropical pixels), activity that increases following rain but decreases with light (28%), and activity that increases following bright seasons but decreases with rain (14%). We compare distributions of these observed relationships with those simulated by land surface models. In general, model simulations of gross primary productivity (GPP) overestimate the extent of positive correlations of photosynthetic activity with water and underestimate positive correlations with light. The largest discrepancies between simulations and observations are in the representation of the regions where photosynthetic activity increases with light and decreases with rain. Our clear scheme for representing the relationship between climate and photosynthetic activity can be used to benchmark tropical seasonality of GPP in land models.