Abstract
SUMMARY: Alternate row mixtures of millet, maize and sorghum were compared with equivalent areas of sole crops. Yield of mixtures were rarely reduced but there was a gain when components differed by a certain margin in height or age at maturity, or both. Approximately 75% of variation in yield gain from mixing was attributable to these two factors. Subsistence farmers who mix a fast-growing early-maturing cereal with a slower-growing, later-maturing one thus reduce interference between neighbouring plants during the reproductive phase of growth, and obtain grain harvest spread in time and heavier than from equivalent areas of the component sole crops.