Crop residues, consisting of cereal stovers and hays made from the vegetative parts of grain legumes, are important dry-season feeds for ruminant livestock in the Savannah Zone of West Africa (van Raay, 1975;McCown, Haaland & de Haan, 1979;Jahnke, 1982). With the increasing demand on land to produce more crops and animal products, traditional systems are under pressure. There is need for a reliable means to quantify the supply of feed from various sources and to monitor the type, quantity and quality of crop residue available to livestock.Various methods have been used to measure crop residue dry matter (D.M.). 'Harvest index' (HI), or the grain/total above-ground D.M. ratio, is often used to determine the grain-producing efficiency of a crop or crop variety. It provides an indirect approximation of the total stover D.M. available to livestock (Dicko, 1980; Okaiyeto & Egharevba, 1981). However, HI are known to vary considerably from crop to crop, between varieties of the same crop, and can be affected by adverse climatic conditions. Also, a HI gives no quantitative value of stover plant parts. The objectives of this study were to test the stability of HI relationships for major crops cultivated in West Africa over a range of management levels as an indicator of the quantity of crop residue on offer and to divide the residue of each crop into component plant parts to gain preliminary information on the relative feeding value of total crop residue D.M.