Electricity generation is an important source of total CO(2) emissions, which in turn have been found to relate to an acceleration of global warming. Given that many OECD countries have to replace substantial portions of their electricity-generating capacity over the next 10-20 years, investment decisions today will determine the CO(2)-intensity of the future energy mix. But by what type of power plants will old (mostly fossil-fuel-fired) capacity be replaced? Given that modern, less carbon-intensive technologies are still expensive but can be expected to undergo improvements due to technical change in the near future, they may become more attractive, especially if fossil fuel price volatility makes traditional technologies more risky. At the same time, technological progress is an inherently uncertain process itself. In this paper, we use a real options model with stochastic technical change and stochastic fossil fuel prices in order to investigate their impact on replacement investment decisions in the electricity sector. We find that the uncertainty associated with the technological progress of renewable energy technologies leads to a postponement of investment. Even the simultaneous inclusion of stochastic fossil fuel prices in the same model does not make renewable energy competitive compared to fossil-fuel-fired technology in the short run based on the data used. This implies that policymakers have to intervene if renewable energy is supposed to get diffused more quickly. Otherwise, old fossil-fuel-fired equipment will be refurbished or replaced by fossil-fuel-fired capacity again, which enforces the lock-in of the current system into unsustainable electricity generation. (C) 2009 Elsevier Ltd. All rights reserved.