Economic and environmental considerations are driving the need for a reliable supply of home grown high protein non-grass forage for organic livestock production. A review of literature highlighted the potential of three lupin species as mono-crops or bi-crops with spring sown cereals, for this purpose. The Conductance model was postulated as a potential tool for predicting the yields and protein contents of such crops and for selecting seeding rates for optimizing protein and dry matter yield. Field trials were carried out in 2005 and 2006 at the Royal Agricultural College, Cirencester, Gloucestershire, UK, to determine the yields, quality and suitability for ensiling of three species of forage lupins grown as bi-crops with a range of spring sown cereals, under organic farming conditions. The Conductance model informed by data from spaced plants was developed as a tool for selecting the most appropriate cultivars and predicting crop performance. Predictions were also made of forage dry matter yields and crude protein contents of the forages in mono and bi-cropping systems. Bi-crop combinations which promised the best potential for yields and quality from the predictions were tested in field trials and the results compared with the predictions. The Conductance model proved to be a reasonably robust tool for predicting yielos and protein contents. Attention was drawn to the strengths and weaknesses of the model as a tool for predicting yield of such crops as those studied. Of the species studied, white lupin (Lupinus a/bus) bi-crops of cv. Dieta with spring triticale (Triticoseca/e Wittmack) cv. Logo or spring wheat (Triticum aestivum) cv. Paragon produced the most suitable combinations for organic production in the prevailing conditions. Dry matter yields achieved at 126 -130 days from sowing were within the range 11 - 20 tha-1 • Bi-crop yields were substantially greater than the constituent mono-crops and biological efficiency, estimated as Land Equivalent Ratio and Nitrogen Equivalent Ratio, was enhanced. Ensiling studies confirmed the superior quality of the bi-crops. Sowing in late April and harvesting at about 112 days from sowing produced marginally the best silages with estimated MEs in a range 11.5 -12.8 MJ kg-1 0M and crude proteins of 100 -195 gkg-10M. It was concluded that in the conditions obtaining on the experimental site that ' ...~ . white lupin/cereal bi-crops produced a reliable and predictable source of good quality non-grass forage which could be recommended for organic livestock production in this part of the UK. While the Conductance model produced good predictions of the yield and quality of vegetative growth, and could be further developed for this purpose with commercial applications, the predictions of reproductive growth and senescence were less convincing and needed further development. Further studies could also be usefullyundertaken to optimize seeding densities for optimum yield and crude protein content of specific forage lupin/cereal bi-crops. Larger plot or field assessments in other locations will be necessary together with animal studies in due course, towards commercialization.