The paper develops a model of the structure of oilseed rape production which is used to investigate the diffusion of the crop in England and Wales between 1974 and 1989. Specifically, the estimates of the structural or aggregate adoption model enable changes in the total area of the crop to be attributed to changes in the farm size structure and in the extent and intensity of adoption by farm size. The results provide direct evidence that rank effects influence adoption behaviour and are also consistent with the existence of epidemic type effects due to the spread of information.