Studies were conducted on some soils of Morogoro District to estimate economically optimum phosphorus (P) fertilizer recommendations for maize from soil tests. The studies involved the estimation of maize yield response to added P in pot and field experiments. Maize responded to added P at three sites. At all three sites, the residual effects of added P lasted up to three years suggesting that added P was not strongly fixed in these soils. Yield responses were related to extractable P contents by the Olsen and AB‐DTPA methods through a modification of the Mitscherlich‐Bray equation. Estimated model parameters were incorporated in equations for estimating optimum P fertilizer rates for different cost‐price ratios (p) and marginal rates of return (R). At the p and R values that prevailed in the study area during 1997, recommended P fertilizer rates (PR) could be determined by the equation: PR=68.497–1.191T (Olsen P). Calculated P fertilizer rates for P deficient sites in the district ranged from 15 to 60 kg P ha‐1. This wide range underscored the importance of site specific fertilizer recommendations.