The lesion nematode Pratylenchus penetrans is a common pest of corn in the north-central United States. There are relatively few studies documenting the impact of Pratylenchus spp. on grain yield even though they are recognized as pests of corn and the target of commercial seed treatments. We adapted a component error modeling approach to develop a damage function for P. penetrans that included the influence of year and site in the yield loss relationship. Field data from six site-years was used to derive panel data consisting of all pairwise comparisons of the difference in nematode population densities and the associated proportional yield difference. Fourteen regression models of the relationship between proportional yield loss and the difference in nematode density were developed from soil and root assays at different corn growth stages. Seven models were significant: four models based on nematode population densities in soil (initial and final samples) and three based on nematode densities in seminal roots (corn growth stages V1 to V2 and V6) and adventitious roots (corn growth stage R1 to R2). The model we consider to be the most important, that based on the initial soil assay, estimated the yield loss caused by each nematode to be 0.0142%. The grand mean of the 118 plots we sampled implied a yield loss of 3.79%. The random effects of year and field did not contribute significantly to any of the models but were close to significance for some, suggesting a benefit from larger data sets. Experimental error was the largest component of the variance for all of the models; therefore, the damage function is more useful for demonstrating impact of P. penetrans rather than for accurately predicting yield loss at the field level. All of the fields in our study were an irrigated loamy sand soil, with grain yields above the county average; therefore, it is possible that our damage function is conservative. The value of soil sampling has been questioned for P. penetrans and this study shows it to be equal to if not better than root assays for predicting yield.