Loss of fixed, filter‐retainable 14C due to treatment with preservatives is reported in productivity studies of natural phytoplankton populations. Treatment with formaldehyde consistently results in large decreases in activity, whereas iodine and gluteraldehyde treatments cause losses with some populations but not with others. The extent of loss with all treatments varies over time in natural populations, suggesting that no simple correction factor can be applied for all populations and that leakage may not be uniform for different species. Should activity losses vary by species, then quantitative autoadiographic methods for estimating individual species production rates may require individual correction factors for each species when preservative‐caused losses occur.