Rapid industrialization and population explosion has led to the migration of people from villages to towns, which generate thousands of tons of MSW daily. One of the main functions of urban local bodies in developing countries is Municipal Solid Waste Management (MSWM). Forecasting of solid waste generation is a vital component of MSWM. This paper deals with modelling of solid waste generation in medium scale towns located in a developing country. Primary data is analyzed to identify independent variables to be used to predict solid waste generation. Different empirical models are used for population projection. Projected population is used to predict solid waste to be handled during the study period. Artificial neural network models are proposed for estimation of solid waste generation in medium scale towns located in the study area. Efficacy of the proposed modelling scheme for forecasting solid waste generation was successfully demonstrated with the help of validation data.