The effect of egg density on establishment and dispersal of larvae of the western corn rootworm, Diabrotica virgifera virgifera LeConte, was evaluated in a 3-yr field study. Implications of these data for resistance management plans for Bt crops are discussed. Viable egg levels of 100, 200, 400, 800, and 1600 eggs per infested plant were evaluated in 2000, 2001, and 2002. A 3200 viable egg level was also tested in 2001 and 2002. All eggs were infested on one plant per subplot in a field that was planted to soybean, Glycine max (L.), in the previous year. For each subplot, the infested plant, three plants down the row, the closest plant in the adjacent row of the plot, and a control plant at least 1.5 m from any infested plant (six plants total) were sampled. In 2000, there were five sample dates between egg hatch and pupation, and in 2001 and 2002, there were six sample dates. On each sample date, four replications of each egg density were sampled for both larval recovery and plant damage. Initial establishment on a corn plant seemed to not be density-dependent because a similar percentage of larvae was recovered from all infestation rates. Plant damage and, secondarily, subsequent postestablishment larval movement were density-dependent. Very little damage and postestablishment movement occurred at lower infestation levels, but significant damage and movement occurred at higher infestation rates. Movement generally occurred at a similar time as significant plant damage and not at initial establishment, so timing of movement seemed to be motivated by available food resources rather than crowding. At the highest infestation level in 2001, significant movement three plants down the row and across the 0.76 m row was detected, perhaps impacting refuge strategies for transgenic corn.