Coastal populations of Macrobrachium ohione (Smith, 1874) have been shown previously to be amphidromous, i.e., with adults living in fresh water but with marine larval development. Larval delivery to coastal estuaries in far-upstream populations seemed unlikely because of the distances involved. Therefore, we tested the hypothesis of freshwater larval development in far-upstream populations from the Mississippi River, near Vicksburg and Greenville, Mississippi. We compared the molting success of newly hatched stage-1 (non-feeding) to stage-2 (first feeding) zoeae when exposed to fresh water and salt water (15 ppt) treatments. In addition we also tested the duration of time larvae spent in fresh water or salt water prior to major larval mortality or larval molting. In all freshwater treatments, stage-1 larvae failed to molt to stage 2; in contrast, molting success in saltwater treatments was ∼99% and after 5-6 d of exposure to salt water all surviving larvae molted to stage 2. In freshwater treatments, there was a significant decline in larval survivorship after 3-5 of exposure to fresh water. Larval survivorship declined below 50% after 5 d of "freshwater drifting." These results suggest that far-upstream populations of M. ohione require saline environments to complete larval development. Alternate hypotheses (long-distance hatching migrations of females to and from the sea, inland brine springs producing low salinity larval nurseries, upstream population sinks) are proposed to explain the former existence of far-upstream populations.