This study examines the effect of direct normal irradiance (DNI) forecast accuracy on the financial value of a concentrated solar thermal (CST) plant. Other factors such as electricity market regulations, plant site local climate, and operating strategy are not considered. A CST model varied over 11 combinations of solar field sizes and storage sizes is used to simulate plant operation for three forecast methods and a perfect forecast. The financial value is calculated using revenue and reserve generation payments resultant from plant operation. Results show when the root mean square error (RMSE) of a 48-h DNI forecast is between 325 and 400 W/m(2), a 1 W/m(2) improvement increases the financial value by $400-1300 per 6 months operation for a CST plant with solar multiple between 1.25 and 2, and storage size between 0 and 20 h. Similarly, when the mean absolute error (MAE) is between 250 and 300 W/m(2), a 1 W/m(2) improvement shows an increase of $1000-3600 per 6 months operation. If two forecast methods have similar MAE or RMSE, then the method that tends to over-predict DNI achieves higher value. For all forecast methods, increasing solar multiple or storage size increases financial value. Financial value expressed using only revenue is overstated by 14-64% compared to using both revenue and reserve generation payments, depending on the CST plant configuration and the forecast method. CST plants with small solar fields or small storage sizes gain proportionally more from investing to obtain better DNI forecasts because more accurate forecasts help these CST plants generate more electricity from the limited solar field thermal output, and use more of the limited stored thermal energy to increase revenue instead of reduce reserve generation payments caused by forecast errors. (C) 2015 Elsevier Ltd. All rights reserved.