The conventional approach of using modern values of lake hydrochemistry and sediment composition as key parameters for identifying proxies for environmental reconstruction is complicated by regional climate, catchment geology, lake morphometry and anthropogenic input. In contrast to modern studies from lakes in the north‐temperate regions, reliable proxy indicators for past environmental change from tropical lake systems are still lacking. In this study, we compare new surface geochemical and mineralogical data from the Manasbal Lake, Kashmir (India), with other lakes situated in climatically and geomorphologically distinct regimes with varying degrees of pollution (Tso Moriri in north‐western Himalaya and Lonar in central India). We characterise external (e.g., geology and weathering), internal (e.g., lake water mixing and evaporation) factors and human activities influencing sediment composition and spatial distribution of proxies in each of these lakes to evaluate the usefulness of some of the commonly used proxies, as well as ascertain conditions where they might be applicable. Our results show that (i) morphometric parameters (such as lake size, water depth and catchment vs. lake area) variably influence the spatial distribution of sediments and proxies on the lake bottom highlighting the limitations of using single cores for long‐term palaeoclimate reconstruction; (ii) hydrochemical proxies (such as Na+/Na++Ca2+ and Mg/Ca) can delineate endogenic (Lonar and Tso Moriri lakes) versus allogenic (Manasbal Lake) influence on modern sediment characteristics; (iii) the compilation of heavy metal data (e.g., Fe, Zn, Mn, Cr and Ni) from all three lake basins with varying degrees of pollution highlights the limitations of using them to characterise weathering conditions in these basins; and (iv) human impact must be precluded before using the most promising endogenic minerals (calcite, aragonite, gypsum and pyrite) as palaeoenvironmental/palaeoclimate indicators.