A previous study in 1968 on this long-term experiment reported the cumulative effects of repeated applications of phosphorus (P) and found that residual effects of applied P resulted in increasing grain yields. A re-analysis of most of that dataset (1943–59) confirmed these results over six cycles of a 3-year rotation (fallow–wheat–oats) with superphosphate applied at 0, 3, 6, 9, and 12 kg P/ha. However, continuation of the experiment in the same plots with a 2-year rotation (fallow–wheat or pasture–wheat) found no further cumulative benefit of residual P on grain yields. Over four decades from 1962 to 2001, yields remained generally constant, after adjusting for growing-season rainfall, with the differences in yield between rates of applied superphosphate also remaining constant over the period. On average, adding 3 kg P/ha was estimated to increase yield by 0.45 t/ha (compared with 0 kg P/ha) and adding 6, 9, and 12 kg P/ha resulted in estimated step-wise yield increases of 0.10 t/ha, 0.10 t/ha, and 0.01 t/ha, respectively.