In Tanzania, and in East Africa in general, bioassessment methods for monitoring and assessing riverine ecosystems are not yet in place. This thesis describes the development of a macroinvertebrate-based bioassessment method for assessing the degree of anthropogenic disturbance in Tanzanian rivers. The hypotheses that, 'macroinvertebrate assemblages reflect disturbance in river systems'; 'rivers with similar abiotic features have similar macroinvertebrate assemblages'; and 'spatio-temporal variation in macroinvertebrate assemblages influence bioassessment', are tested. Macroinvertebrate and environmental data were collected from the Pangani, Rufiji and Wami-Ruvu basins. Univariate analyses; constrained and un-constrained ordinations and a linear response model were used to test the hypotheses. Five important bioassessment aspects were investigated. A set of 20 criteria for screening reference sites was established and used to identify and distinguish between reference and test sites in the study area. A two-level hierarchical framework for classifying homogenous river types was developed and validated. Three river types were classified: Pangani highland uplands, central eastern Africa uplands and central eastern Africa lowlands, each with two sub-Groups. A macroinvertebrate-based biotic index, the Tanzanian River Scoring System (TARISS), was established for monitoring and assessing anthropogenically induced disturbance in Tanzanian rivers. TARISS has three metrics; number of taxa, TARISS score and average score per taxon (ASPT) for measuring disturbance. Spatio-temporal variations in macroinvertebrate taxa, assemblages and TARISS metrics were examined. Spatial variation within river types was driven by catchment characteristics such as geographical location, geology, altitude and local characteristics such as active channel width, proportions of boulder, cobble and sand on the bottom, influenced reference conditions in all three river types. Temporal variations were significant in the central eastern Africa lowlands, with higher TARISS metrics in wet than in dry periods. Biological and physico-chemical reference conditions were identified for each river type and sub-Group. Guidelines for interpreting TARISS data were established for the validated sub-Groups. In conclusion, TARISS proven to be reliable in detecting anthropogenic disturbance in Tanzanian rivers and is recommended as a national bioassessment method.