Experiments at seven locations in Western Canada between 1956 and 1961 indicated that three or four tillage operations were usually enough for satisfactory weed control on summerfallow, and generally provided highest yields of grain. Two tillage operations were not sufficient. The herbicides that were available did not control all of the weeds that were present. Nitrate nitrogen accumulation at the end of the summerfallow period was greater when weeds were adequately controlled by tillage than with incomplete control. Different amounts of tillage had little effect on soil-moisture conservation. The use of herbicides without tillage resulted in inadequate weed control in most years, and slightly less moisture was stored.