A technique was developed for determining the sediment oxygen demand (SOD) and iron and manganese release rates under laboratory conditions. Site-specific comparisons of this method with the EPAin situtechnique, indicated a consistently lower SOD rate but improved reliability. Manganese release began as the dissolved oxygen concentration approached 1 mg/L and below. Iron release did not begin until –100 to –150 mv oxidation reduction potential was reached. A high degree of consistency was indicated between replicates. Laboratory andin situSOD rates yielded a two-dimensional model of reservoir oxygen budgets. The laboratory rate compared favorably to the assumed SOD rate of the model. As a result, TVA has incorporated the laboratory-determined rates and iron and manganese release rates into the reservoir management planning process. Specific measured and modeled rates determine waste load allocations for several reservoirs.