Planning and engineering activities in small-scale power systems are, in most of cases, driven by immediately pressing factors such as short-term demand, system costs, expected revenues, and local development priorities. Often, the decision to go ahead with the investment in such systems is based on the outcomes of single-attribute spreadsheet type analyses and linear optimization runs where key parameters such as demand growth, interest rates, capital costs, and fuel prices are assumed to remain constant throughout the study period. The least-cost plan thus obtained is subject to changes in the above parameters requiring continual re-evaluation and assessment to bring the project up to date. An alternative method is hereby presented in which the uncertain parameters in a seemingly deterministic model were identified ahead of time. The range of values that each of these parameters could assume as well as the respective probabilities were elicited from the experts and incorporated in a decision analysis problem designed to generate the least-cost policy. The decision analysis process resulted in a robust evaluation of generation options under investigation when compared to the results of the deterministic analysis. Moreover, options ranked least in the deterministic analysis became quite competitive when uncertainties were included in the analysis. / Master of Science