Postincubation differential filtration (PIDF), preincubation differential filtration (Pre-IDF), and track autoradiography (TA) were compared for estimating cell-specific and total photoautotrophic picoplankton production. Experiments were performed in Lakes Michigan and Huron and in the Gulf of Mexico. When Synechococcus dominated the photoautotrophic picoplankton community (> 70% of total picoplankton abundance), PIDF estimates of cell-specific and total picoplankton production were approximately 3.0 x (range, 2.0-3.8 x ) higher than TA estimates. PreIDF estimates of cell-specific and total picoplankton production, however, were only slightly higher than TA estimates (mean, 1.4 x ; range, 1.4-1.5 x). The higher PIDF estimates were attributable to breakage and damage of larger photoautotrophs during postincubation filtration and to retention of this labeled material on the smaller (0.2 mum) pore-size filter. Results from PIDF experiments must be viewed with caution and previous estimates of picoplankton production, cell-specific or total, based solely on PIDF may need to be re-evaluated.