Maize (hybrid PX61O) was sown at 85000 (Treatment 1), 181000 (Treatment 2), and 362 000 (Treatment 3) seeds per hectare in large plots (1.0 ha in 1975-76; 0.5 ha in 1976--77). In 1976--77 (Year 2) deaths of maize plants were greatest during the first 110 days growth and amounted to 10, 11, and 24% of established seedlings for Treatments 1, 2, and 3, respectively. Total dry weight per plant, the proportions of leaves, stem, grain, cob, and tassel, and the concentrations ofN, P, K, Na, Ca, and Mg, were assessed at intervals until the 'hard dent' growth stage (30-35% total plant DM), about 160 days from sowing. High plant population densities had little effect on mineral concentration, but reduced total dry weight per plant and reduced the weight of individual shoot components, especially leaves, stems, and grain. For Treatments 1,2, and 3 respectively in Year 2, grain represented 38,28, and 21% of plant dry weight at 144 days from planting and 49, 43, and 33% of plant dry weight at 167 days from planting. At the 'hard-dent' growth stage in 1975-76 (Year 1), sampling estimates of DM yield per hectare of 21.3, 27.4, and 36.1 t.ha-1 for Treatments 1, 2, and 3, respectively, were 12, 38, and 68% greater than yields measured by harvesting whole plots. In Year 2, yield estimates of22.0, 25.0, and 24.4 t.ha-1 of DM for these treatments were 8, 14, and 4%, respectively, above yields measured by total harvest. The optimum plant population for maize silage, the growth and development of maize grown at different plant populations, the quality of the crop at ensiling, and problems encountered in accurately sampling large plots of maize for estimation ofDM yield per hectare are discussed.