Crop establishment is a major concern for producers who grow legumes on sandy soils. This study was conducted on a Verndale sandy loam (coarse-loamy mixed Udic Argiboroll) to evaluate the impact of five methods of establishment on the production of four legumes (alfalfa [Medicago sativa L.], birdsfoot trefoil [Lotus corniculatus L.], cicer milkvetch [Astragalus cicer L.], red clover [Trifolium pratense L.]). Methods varied from the use of a clean, prepared seedbed to no-till seeding into small grain stubble. The study was initiated in 1991 and repeated in 1992. Yields from spring seedings were measured in the year of establishment. Dry matter yields were also measured for two resided years following establishment. Stand counts were taken in the fall of 1992. Yields of all legumes measured in the year of establishment were generally higher when there was no competition from a companion crop. In general, yields of alfalfa and red clover were equal but higher than yields of birdsfoot trefoil and elect milkvetch when all establishment methods are considered. Stand counts showed that all establishment methods produced an acceptable stand of all legumes. Measured stand, however, was not directly related to forage yield. The results of this study show that there are several methods that can be used to successfully establish legumes on sandy soils. Choices for individual forage producers will be highly dependent on erosion potential and crops grown for the total farming system.