Forecasting crop production is an important dimension when monitoring agricultural markets. Over time, forecasting techniques have evolved, as has agriculture itself and the specifications of the forecasts that are needed. Those who use forecasts seek greater accuracy, granularity, comparability, and timeliness; those who produce the data frequently face financial and technical constraints. Obtaining timely knowledge presents a very real challenge. The human, institutional, technical and financial infrastructure needed to produce crop forecasts is complex. This publication provides insights into these complexities, highlighting good practices from selected country case studies and outlining prospects for the future. By describing the functioning of diverse institutional [...]