Seven experiments which included one field/agronomy, four in vitro and two in vivo trials were carried out to (i) determine the optimum time of harvesting and storage methods that preserved the nutritive quality of maize and sorghum stovers, (ii) examine strategies for enhancing the utilization of high tannin feeds (bird-resistant (BR) sorghum stover supplemented with forage legumes (FLs) with varying tannin concentrations) by using ameliorants (polyethylene glycol MW 4000 (PEG-4000), urea or sulphur). It was determined that nutritive value was significantly (P < 0.001) influenced by crop type, period of harvest and storage method. The highest nutrient content was recorded at 2 weeks (-2 weeks) before physiological dead ripe stage of grain and when the stover was harvested immediately after grain harvest and stored under shade. Addition of ameliorants significantly improved DMD and gas production in vitro. Urea and sulphur were as effective as PEG in alleviating the anti-nutritive effects of high tannin in feeds. The optimum levels of PEG required to alleviate the anti-nutritive effects of CTs varied with forages. Nevertheless, a ratio of 2:1 (PEG: CT) was found to be optimum for the alleviation of the adverse effects of CTs. Forage legumes with high concentrations of CTs, when used as supplements to BR sorghum stover, depressed animal performance (feed intake, growth and growth efficiency). The response assumed both a linear and quadratic pattern, suggesting a threshold point beyond which response decreased. Mixing FL ('Lablab purpureus', 'Chamaecytisus palmensis', ' Sesbania goetzei 15007', 'Sesbania sesban 15019', ' Desmodium intortum' and 'Acacia angustissima') with different nutritional attributes at 1:2 or 2:1 ratio significantly improved gas and ammonia production in vitro. Thus there is potential for improving the utilization of some forages that would normally pose problems (e.g. ' A. angustissima') if used by themselves. It was possible to assess the reactivity (biological activity) of tannins by reacting them with PEG and relating it to radial diffusion assay (protein precipitation assay). Results showed a high correlation (r = 0.87, P < 0.01) between PEG-CT reactions and radial diffusion assay, thus indicating that PEG-CTs reaction may be an effective method of indexing tannin reactivity.