["Journal of Agricultural Economics, EarlyView. ", "\nABSTRACT\nHow large is the potential benefit for farmers from obtaining forecasts of total rainfall in the upcoming season? This critically depends on the local farm system and the crop management actions available to farmers to respond to the forecasts issued. Studies assessing the financial value of seasonal precipitation forecasts have so far focused almost exclusively on start‐of‐season crop sowing decisions using approaches for non‐embedded risk analysis. However, crop management decisions are spread over the entire growing cycle, and considering this realistically could either increase or diminish the financial value of seasonal forecast information from a theoretical perspective. Our simulation analysis demonstrates how this research gap can be closed: we employ a state‐contingent embedded risk approach combining multi‐stage discrete stochastic programming and high‐performance computing to determine optimal crop management responses (crop choice, sowing date, tied‐ridging, relay cropping and fertilization) to seasonal precipitation forecasts throughout the full crop growth cycle. For rain‐fed agriculture in the Central Rift Valley of Ethiopia, we find seasonal forecasts to increase risk‐adjusted expected cash surplus by about 2%–3% at realistic forecast skill levels. For a typical risk‐averse subsistence farming household (0.5 ha), we calculated an increase in long‐term expected cash surplus of 58 USD PPP at 2023 prices. Intra‐seasonal crop management plays an important role in forecast response. A non‐embedded risk approach simulated not more than 35% of the forecast's economic value simulated with the embedded risk approach.\n"]