Tests were conducted on Selenastrum capricornutum and lake water to evaluate the common practice of estimating abiotic uptake of radiophosphorus (33Pi) by poisoning samples with glutaraldehyde (GA) and Formalin (FM) and to estimate abiotic uptake in Lake Michigan water. Algae treated with GA and FM release intracellular 31Pi into solution, which elevates the dissolved 31Pi:33Pi ratio and could seriously underestimate abiotic uptake. Carbonyl-cyanide m-chlorophenylhydrazone (CP), an inhibitor of phosphorylation, was identified as a satisfactory agent for estimating abiotic uptake because it effectively inhibits biological uptake, it does not cause the release of detectable amounts of 31Pi from algae into solution, and it does not appear to block significantly abiotic uptake by particulate material in lake water. Two types of tests demonstrated that nonbiological uptake of 33Pi in water from Lake Michigan was negligible: uptake in samples treated with 10−3 mol/L CP was ≤ 0.01%, and uptake by particulate material recovered from lake water after treatment with heat or GA and resuspension in filtered lake water was ≤ 2% of rates measured in untreated samples. Procedures for estimating abiotic uptake in 33Pi tracer experiments are proposed.Key words: methodology, carbonylcyanide m-chlorophenylhydrazone, Lake Michigan, phosphorus, phytoplankton