It has been proposed that bidirectional and concentration-dependent fluxes of digestive enzymes across pancreatic acinar cell membranes account for secretion. One implication of such a model is that protein secretion should be a function of fluid outflow, inasmuch as flow would be required to generate the necessary concentration gradient by carrying away secreted material. In an earlier study (Science 204: 1212, 1979) when fluid flow was decreased by a backpressure applied to fluid in the pancreatic duct, proportional reductions in protein secretion occurred. The present study uses metabolic rather than mechanophysical methods to decrease flow, reduction of the sodium concentration of the medium bathing the pancreas, or addition of the Na+-K+-ATPase inhibitor ouabain. Both treatments produced similar results: decreases in protein output synchronous with and proportional to the observed decreases in flow. Essentially the same relationship was seen when flow was reduced during protein secretion augmented by the secretagogue cholecystokinin-pancreozymin. These results suggest that the reduction in flow rate (whether produced mechanically or chemically) was the variable directly responsible for the decrease in protein secretion by the acinar cell.