Soybeans within a specific seed size range are often preferred for the manufacture of specialty soy products. This study examined the influence of row spacing and seeding rate on seed yield, average seed size, and seed size uniformity of Camp (small seeded) and V71-370 (large seeded) soybean cultivars [Glycine max (L.) Merr.]. Experiments were conducted in 1992 and 1993 at Kentland Farm near Blacksburg, Virginia. Row widths of 38 and 76 cm and seeding rates of 125, 250, 500, and 750 thousand seeds ha⁻¹ were used. Averaged over treatments, 38-cm row spacings yielded higher than 76-cm row spacing by 11.6%. Additionally, each increase in seeding rate resulted in increased yield up to 500,000 seeds ha and then leveled off. Yield of useful seed of Camp (total yield multiplied by the proportion of seed between 4.8 and 5.6 mm in diameter) was maximized at narrow row spacings and 500,000 seeds ha⁻¹. Neither row spacing nor seeding rate influenced V71-370 yield of usable seed (larger than 7.9 mm). Averaged over years and cultivars, soybean seed size was not influenced by row spacings or seeding rates. However, in 1993, smaller Camp seed was favored by narrow rows and low seeding rate. Seed size uniformity, as determined by analysis of variance of individual seed weights was, in general, not influenced by row spacing or seeding rates. In a separate investigation in 1992, heritabilities were estimated by two methods using plot data from two replications of lines from crosses of Camp with Jizuka, MD87L-0198, and Stafford soybean cultivars. Heritabilities based on F₃ progeny means regressed on F₂ parental means ranged from 24% to 41% and were lower than the corresponding heritability estimates based on analysis of variance which ranged from 43% to 83%. Correlations of seed size with other agronomic traits were generally low. Although low in magnitude, significant positive correlations of seed size with yield in some crosses might hinder progress of breeders when selecting for small-seeded, high yielding cultivars.Ph. D.