Two seasons of cropping were carried out at three sites in the Guinea savanna to evaluate the residual effects of soyabean on maize. The experiment was laid out as a splitplot design in a randomized complete block with three replications. In the first season, four soyabean varieties with a fallow treatment (control) received phosphorus (P) applied as triple superphosphate (20% P) at the rates of 30 and 60 kg P ha
−1
. Maize was grown in these plots in the second season without fertilizer application. At all sites, regardless of the previous crop, total soil N remained low (<1.5 g kg
−1
). Available P was affected by the P rate in the previous year at all sites. From initial values ranging from 5.2–16.2 mg kg
−1
in the first season, available P significantly (p<0.05) increased in the second season to 9.8–42.8 mg kg
−1
when 30 or 60 kg P ha
−1
was applied, compared to 7.7–18.6 mg kg
–1
at no P application. Relative to no P application in the previous year, the application of 60 kg P ha
−1
significantly increased total dry matter at 6 weeks after planting by 19%, total harvest dry matter by 28%, and grain yield by 37%.