This study determines the optimal investment portfolio in Tehran Stock Exchange (TSE) industries. For this purpose, a conditional capital asset pricing model (CAPM) with time-varying covariance, according to a Multivariate GARCH approach has been formulated. According to this conditional CAPM, the conditional variance-covariance matrix and mean of returns are calculated for some industries. By using the Mean-Value at Risk portfolio selection model, the optimum proportion is detected. Results showed that the Pharmaceutical Industry, Financial Group and Cement Industry have the most quotas in portfolio since they maintain the minimum variance and maximum return among all other industries.