Red clover (Trifolium pratense L.) seeding associated to a winter wheat (Triticum aestivum L.) is a usual practice among farmers, in the province of Arauco, Chile. The assessment of its effects upon wheat grain yield and pasture productivity and detecting the optimum red clover seeding date were the aims of th is experimento In 1981, in a red-clayed soiI of the Cañete area, red clover was broadcasted in May, June, July and August, over two wheat varieties: Capelle Desprez (normal height) and Manquefen (semidwarf), both sown in May. Other treatments were: clover sown in rows, with both wheats, in May: clover sown alone in rows, in April, May and August; and both wheats sown alone, in May. Seed ing rates were 150 kg/ha for wheat and 12 and 18 kg/ha of clover, for rows and broascasting, respectively. Result indicated no yield differences in wheat, when associated with clover, at any time, in either cultivars. However, grain yields were significantly higher (P < 0.01) for the normal height cultivar (5.1 vs 4.5 Ton/ha). Dry matter yield of clover, at wheat harvest time in the associated sowings, did not overcome 0.5 Ton D.M./ha, and there were not differences (P > 0.05) among treatments. Clover seeded alone in April and May, yielded 2.94 and 1.38 Ton D.M./ha, respectively; clover sown alone in August was statistically similar in yield to the associated treatments. During the second year, clover production decreased in a quadratic form, when planting date was postponed, from 9 to 2.3 Ton D.M./ha, under the semidwarf wheat, and from 7.5 to 0.5 Ton, under the normal height cultivar; the cultivar effect was significant (P < 0.01).