Field studies using area-of-influence techniques were conducted in 1987 and 1988 to evaluate soybean growth and yield, and to predict soybean yield losses from photographs of jimsonweed and common cocklebur canopies. Differences in weed competitiveness within the 100-cm area of influence were induced by dates of soybean planting, locations, weed species, and years. Soybean yield losses within the first 20-cm interval from weeds correlated well with yield of all soybean plants within 100 cm of weeds (r2 = 0.86). Soybean growth responses as a function of distance from weeds were best described by complex polynomials, but simple linear functions, based on a data point from soybean plants nearest a weed and from the average of plants 60 to 100 cm from a weed, closely approximately actual yield losses (r2 = 0.96). Soybean yield losses were highly correlated (r2 = 0.84) with leaf area of weeds as viewed from directly above the weed-crop canopy. Weed canopy diameter, measured from overhead photographs 8 weeks after soybean emergence, also correlated well with soybean yield losses (r2= 0.82), but correlation with actual weed leaf area was not significant (r2 = 0.31).