After a review of hydropower optimization models, we focus on price-driven hydropower dispatch models under uncertainty of the electricity price. We present two modeling approaches for pumped-storage plants. In the first model, the water level is constrained in expectation. We discuss the marginal price of water, which is obtained analytically, and influences of price variances. The second model is a multistage stochastic linear program on a scenario tree. Financial risk is constrained by a time-consistent extension of CVaR (conditional-value-at-risk). The model has two time scales: The short-term dispatch decision is separated from the long-term planning by aggregating electricity prices into occupation times at price levels. The risk constraint is tested in a case study.