This paper proposes an application of the option valuation approach to evaluate a project investment in three stages: research (R), development (D) and acquisition (A). To reflect different correlation and effect of each type of uncertainty on option values and, consequently, investment decisions, the proposed valuation and decision model incorporates both technical and market uncertainties into the first two technical stages (R&D) of the project, and the market uncertainty only into the last stage (A). Changes in project values are accordingly captured in each stage by the combined geometric Brownian motion and Poisson jump downward processes. The model incorporates the patent sale alternatives in the development and acquisition stages. A dynamic programming, decision tree model is solved to determine the option values and optimal decisions subject to decision rules, critical values, and certain boundary conditions. We subsequently evaluate the model effectiveness by comparing its decisions with those of an existing valuation model and the net present value method. The Monte Carlo simulation results show that under the option valuation by which the loss is limited to the initial costs of investment, a positive profit in a wide range can be obtained with more than 50% chance, in spite of the small average profit. The results of simulation also verify the significance of the chance to sell the patent as a safer decision under the high market uncertainty. In addition, a shorter term of patent agreement significantly improved advantage of the patent sale over the immediate investment in an optimistic situation.