Sixteen experiments have investigated alternative methods of predicting the competitive effects of a simulated weed (oats) on the yields of spring barley, spring oilseed rape (canola), peas, spring field (faba) beans and flax. The experiments were designed to discover whether early postemergence assessments of crop and weed vigor would achieve more reliable prediction of yield loss than weed density. Weed density (plants m−2) was a very variable predictor of yield loss. The standardized ranges (range/mean) of values over 3 to 4 years of data for the five crops, in the densities causing 5% yield loss, were between 1.14 and 2.59. Predictions based on the relative dry weight of crop and oats (oat dwt/(oat dwt + crop dwt)), assessed while the plants were still small, achieved more reliable predictions, as the standardized ranges were between 0.10 and 1.86. In three of the experiments, predictions based on relative dry weights were compared to similarly timed predictions based on measurements of relative leaf area and of ground cover, assessed subjectively (by eye) and photographically. Subjective and objective (photographic) assessments of cover achieved similar predictions of yield loss, indicating that visual assessments could be a viable tool to assess the potential competitive effects of weeds.