Phytoplankton play a key role in biogeochemical cycles, impacting atmospheric and aquatic chemistry, food webs, and water quality. However, it remains challenging to reconstruct changes in algal community composition throughout the geologic past, as existing proxies are suitable only for a subset of taxa and/or influenced by degradation. Here, we investigate if compound-specific hydrogen isotope ratios (δ²H values) of common algal lipids can serve as (paleo)ecological indicators. First, we grew 20 species of algae – representing cyanobacteria, diatoms, dinoflagellates, green algae, and cryptomonads – in batch cultures under identical conditions and measured δ²H values of their lipids. Despite identical source water δ²H values, lipid δ²H values ranged from −455 ‰ to −52 ‰, incorporating variability associated with chemical compound classes and taxonomic groups. In particular, green algae synthesized fatty acids with higher δ²H values than other taxa, cyanobacteria synthesized phytol with relatively low δ²H values, and diatoms synthesized sterols with higher δ²H values than other eukaryotes. Second, we assessed how changes in algal community composition can affect net δ²H values of common algal lipids in 20 experimental outdoor ponds, which were manipulated via nutrient loading, and the addition of macrophytes and mussels. High algal biomass in the ponds, which was mainly caused by cyanobacterial and green algal blooms, was associated with higher δ²H values for generic fatty acids, relatively stable δ²H values for phytol and the dinoflagellate biomarker dinostanol, and lower δ²H values for the more cosmopolitan sterol stigmasterol. These results are consistent with expectations from our culture-based analyses, with both datasets indicating large taxon-specific changes that are unlikely to be driven by bacterial heterotrophy. This suggests that measuring δ²H values of multiple lipids from sediment and calculating ²H-offsets between them can resolve changes in algal community composition from changes in source water isotopes. With an appropriate availability of sedimentary lipids, this approach could permit the reconstruction of both taxonomic variability and hydroclimate from diverse sedimentary systems.