Monte-Carlo simulation was used in a capital budgeting decision involving the purchase of a wood-fired cogeneration plant. Two alternative capacities, 6-MW and 9-MVV plants, were considered in the analysis using net present value as the evaluation criterion. Risk profiles were generated for both investments as well as for the mean and variance of returns. When cash flows were approximated using annual cash flows, the resulting risk profiles did not produce a clear choice. However, using more realistic monthly cash flows, the 6-MW plant showed both a higher return and less risk than the 9-MW facility. The difference in the results arises from the fact that the annual-period model masks the effects of the inherent monthly variability of some cash flows and timing of cash savings. Consequently, the use of probabilistic, monthly-period models is recommended over the annual-period model in this as well as in other investments that incorporate a large degree of seasonal variability.