The relative leaf area of weeds is a good predictor of the outcome of weed-crop competition. However, this variable has not been used in decision-making tools for integrated weed management because leaf area cannot be measured quickly. A powerful image analysis system for measuring leaf cover (the vertical projection of plant canopy on the ground) has been developed and validated. This research was conducted to compare the efficiency of weed relative leaf area and relative leaf cover in predicting corn yield loss. Field studies were conducted in 1996 and 1997 using varying densities of common lambsquarters, barnyardgrass, common lambsquarters plus barnyardgrass, and a natural weed community. Corn grain yield and biomass loss varied with weed infestation type and year. Values of the relative damage coefficient of weeds (q) were smaller in 1997 compared with 1996. For both years, the relative leaf area of weeds was an adequate predictor of corn yield loss (r
2 varied from 0.61 to 0.92). The precision of the predictions was not influenced by the leaf area sampling period (four- or eight-leaf stage of corn). In general, smaller values of q and m (predicted maximum yield loss) were obtained as a consequence of using the relative leaf cover of weeds in model fitting. However, percentages of variation explained by the model (from 0.67 to 0.90) were similar to values obtained with the relative leaf area. On the basis of the residual mean squares, neither of the variables could be declared superior to the other in yield loss prediction. The development of weed control decision-making tools using the relative leaf cover of weeds may require improvements prior to being used in weed management systems. Such improvements would include use of appropriate sampling and image-processing techniques, development and validation of empirical models specific to individual situations, and proper identification of the crop growth stage at which leaf cover must be assessed.