In the context of decaying capital cost and uncertain revenues, prospective valuation of a wind power distributed generation (DG) project is difficult. The conventional net present value (NPV) presents a static picture that does not account for the value of waiting for better market conditions to proceed with a DG investment. On the contrary, real options (RO) analysis does account for the managerial flexibility to switch between options over the investment horizon. In this paper we argue that the value of a DG wind-based project can be revisited by means of Longstaff-Schwartz method, originally intended for the evaluation of American financial options. The adaption of this method to the wind power DG setting provides a means for (i) efficiently dealing with the several stochastic processes involved (spot electricity prices and possibly various wind speed processes) avoiding the curse of dimensionality, (ii) accounting for the decaying capital cost of DG, and (iii) solving the perfect foresight problem presented by Monte Carlo conventional simulations. We present in this paper the procedure to follow when applying the method to the wind power DG setting. Particularly, we discuss the standardization of the wind speed and spot price processes, and the advantages of building a state space model that includes all the correlated processes by adequately transforming Box-Jenkins and Ornstein-Uhlenbeck models. Also we discuss the representation of the capital cost forecast by means of learning curves. On the whole, we put together this multivariate setting and show how it is translated to the Longstaff-Schwartz framework. The developed methodological approach allows us analyzing the sensitivity of the RO value to the site and DG capacity factor, the risk-free rate, the current forward projections of global installed capacity, and the aggregation of DGs under a common owner. On the whole, we conclude that the RO analysis is best employed to analyze DG settings in which the NPV yields a close to negative assessment: low power production, for instance. In such cases, RO analysis shows advantages in producing delayed optimal investment times. (C) 2015 Elsevier Ltd. All rights reserved.