Soil fertility remains a persistent constraint for smallholder farmers in northern Ghana. The purpose of this study was to quantify the effects of P and N fertilization combined with pigeon pea (Cajanus cajan) and groundnut (Arachis hypogea) rotations on soil fertility and subsequent maize yields. An experiment was conducted in 2009 and 2010 in two villages Piisi and Nakor in Upper West Region of Ghana in collaboration with local farmers and researchers from the Savanna Agricultural Research Institute. A 4x2 factorial split-plot design was replicated for 12 farm households in the area. Main plot treatments were crop sequences of maize-maize (MZ10), maize-maize+60 kg N (MZ60), groundnut-maize (GN), and pigeon pea-maize (PP).  Sub-plot treatments of 0 and 26 kg P ha^-1 were applied as triple super phosphate in 2009 at planting. Nitrogen was applied as ammonium sulfate to all maize plantings at a rate of 10 kg N ha^-1 except for the second year of maize when one of the maize-maize sequences received 10 kg N ha^-1 (MZ10) and one received 60 kg N ha^-1 (MZ60). Measurements of total maize dry matter were taken at harvest in October of 2010. Total dry matter ranged from 6790 kg ha^-1 for MZ10 to 8900 kg ha^-1 for GN with a significant difference between GN and MZ10 and a slightly significant difference between PP and MZ10. Leaf dry matter ranged from 590 to 700 kg ha^-1 and stem dry matter ranged from 1060 to 1220 kg ha^-1, but neither showed significant treatment effects. Maize grain yield was 2420 kg ha^-1 for GN, followed by 2280, 2140, and 1770 kg ha^-1 for PP, MZ60, and MZ10. No effects of phosphorus application on subsequent maize yield were observed. No effects on soil N or soil organic matter were observed. The higher yields of GN and PP as compared to MZ60 despite no significant increase in soil N suggests benefits of rotation beyond soil fertility.