The loss of natural land-cover is one of the main drivers of biodiversity change in terrestrial ecosystems. Driven by the increasing global demand for agricultural commodities, deforestation remains widespread and rapid in the tropics, which also supports a disproportionately large number of species. Generalizing the responses of tropical freshwater biodiversity to these land transformations remains a daunting challenge, as evidence has appeared contradictory. Therefore, we performed a systematic literature search on Web of Science and Scopus with 56 keywords to identify published studies examining tropical fish diversity across contrasting categorical land covers (e.g., intact forests vs. croplands vs. mining areas) and continuous gradients of forest cover (e.g., % riparian deforestation). After screening 1403 references from the literature search, we have identified 250 studies (94 categorical studies and 156 gradient studies) from 45 countries spread across all four major tropical regions: 64.8% of the studies were in the Neotropics, 18% in Indomalaya, 8.6% in the Afrotropics and 4.8% in Australasia and Oceania). We partnered among dozens of authors from these studies (the FISHLUC working group1) to build a pantropical database comprising currently 136 datasets with a total of 5049 site-level fish communities with approximately 2918 valid species. We performed a meta-regression analysis, which revealed the importance of functional identity, land-cover type and intensity, and sampling design in driving the heterogeneity of responses of taxonomic and functional fish diversity to tropical land-cover and land-use change (LCLUC). Last, we will provide this database as a publicly available resource that includes geographic coordinates, species abundances, sampling methods, and standardized LCLUC metrics associated with the fish sampling sites. Our database and results will serve to improve conservation and restoration actions and priorities aiming to mitigate the impact of LCLUC on freshwater ecosystems.