This paper focuses on the application of risk-based models to cogeneration system planning in the presence of uncertainties. Starting from the electrical and thermal load patterns, and the electricity and gas prices for a reference year, the uncertain variables for the successive years of analysis (demand and energy prices) are represented by using average values and covariance matrices, taking into account their possible correlations. A multi-year planning analysis is carried out by means of Monte Carlo simulations, calculating the probability distributions of the net present costs. The same procedure is repeated by considering the base case of separate production and a number of alternatives using internal combustion engines of different size. Post-processing of the net present costs is run according to risk-based concepts, by formulating a composite objective function that takes into account a weighted sum of the average value and of a risk-based component. Hence, instead of evaluating only average cases or the limit cases (best or worst cases), the results obtained from the risk-based calculations provide useful indications for the decision makers, taking into account different degrees of risk attitude of the decision maker.