Seasonal estimates of metabolism were determined with closed, replicate, 12-liter recirculating chambers on an artificially constructed gravel bar in a seventh order stream, the Tombigbee River, and on a natural riffle in a fourth order stream, the Buttahatchie River. In the Tombigbee and Buttahatchie rivers, the reypettive ranges for metabolism parameters were as follows (in g 0 m d : gross community 2 primary productivity, 0.39 to 2.97 and 0.06 to 0.44; net community primary productivity, -0.72 to 2.38 and -0.50 to 0.11; community respiration, 0.60 to 6.00 and 0.24 to 1.04; net daily metabolism, -3.22 to 1.61 and -0.90 to -0.01; and ratios of gross community primary I productivity to 24 hour community respiration (P/R), 0.38 to 2.01 and 0.09 to 0.98. Stepwise multiple linear regression indicated that most of the variation in metabolism parameters was explained by temperature, detritus, substrate-level light, and total Kjeldahl nitrogen in the Buttahatchie River, and by turbidity, temperature, chlorophyll, and nitrate-nitrogen in the Tombigbee River. Both rivers were predominately heterotrophic. The Tombigbee River was autotrophic only in January, at which time an algal-based community existed. During all other sampling periods, both communities were dominated by bacteria and utilized allochthonous inputs.