Optimum crop production depends, among other things, on the maintenance of adequate plant nutrients in the root zone. The objective of this study was to find a reliable index for assessing needs for supplemental phosphorus (P) in soils of Morogoro District, Tanzania. Six indices of P availability, namely: Bray and Kurtz No. 1 (BK1), Bray and KurtzNo.2 (BK2), Mehlich 1, Mehlich 3, Olsen and ammonium bicarbonate‐DTPA (AB‐DTPA), were evaluated. Evaluation of the P indices involved relating extractable P contents by different methods with crop response data expressed as relative yields. The response data was obtained from pot trials with soil samples from ten repesentative soils designated as benchmark soils of the district. Treatments were absolute control, 0, 10, 20, and 30 mg P kg‐1of soil. Correlation of maize relative yields with soil test values by the six indices of P availability resulted in correlation coefficients ranging from 0.65 to 0.90. The Olsen method gave the highest r value suggesting that it was superior to the others. However, using the Cate and Nelson approach, the Olsen and ammonium bicarbonate‐DTPA methods were found to be atparand superior to the others. They each accounted for 76% of the variations observed in maize relative yields, respectively. The critical P levels for the indices were 10.50 mg P kg‐1for Olsen and 2.80 mg P kg‐1for the AB‐DTPA method. Phosphorus fertility categories were delineated in relation to Olsen extractable P as: low (<6.50 mg P kg‐1), medium (6.50 to 23.0 mg P kg‐1), and high (>23.0 mg P kg‐1). Based on this classification it was determined that 16%, 25 %, and 59% of the surveyed area had low, medium, and high P levels, respectively. About 40% of the surveyed area may, therefore, require fertilization with P for optimum yields.