This work summarizes the measurements and associated environmental dosimetry of reactor-released 137Cs and 134Cs and weapons-produced 137Cs in samples of water, shoreline sediment and fish collected from 1971 to 1980 in the Hudson River Estuary. Trends observed in annual mean concentrations and the resultant dose implications for man from each source are discussed. The human exposure pathways examined are: fish consumption, water consumption, swimming and recreational use of the shoreline. Based on environmental measurements, a maximum, adult, whole-body, 50-y committed effective dose equivalent (CEDE) of 0.79 mu Sv (79 mu rem) is estimated from fish consumption in 1971, the year of maximum reactor discharge of the radiocesiums. For comparison, during the period 1974-79, mean estimates (+/- 1 SD) of the CEDE based on environmental measurements and attributed to other pathways are as follows: consumption of indigenous fish species caught downstream of the reactor outfall, 0.05 +/- 0.02 mu Sv (5 +/- 2 mu rem); consumption of fresh water sampled upstream of the reactors, 0.02 +/- 0.03 mu Sv (2 +/- 3 mu rem); and swimming, 10(-4) +/- 10(-4) mu Sv (0.01 +/- 0.01 mu rem). In addition, external, whole-body exposure resulting from recreational use of the shoreline 1.6 km downstream of the reactors is estimated to be 1.2 X 10(-8) C kg-1 (46 +/- 11 mu R yr-1). The above dose estimates are based on consumption factors of 3.9 and 803 kg y-1 (fish and water, respectively) and on usage factors of 50 and 140 h y-1 (swimming and shoreline recreation, respectively). Differences in dose estimates obtained from these long-term environmental measurements and from assessment models currently recommended for use by the U.S. Nuclear Regulatory Commission (NRC) are discussed.