There are several indications that legumes are capable of accessing sparingly soluble phosphorus (P) in the soil through root-induced processes. We hypothesize that this plant-induced mobilization of P can be demonstrated if the plant accessible P assessed by isotopic dilution ('L-value') exceeds the corresponding values assessed in soil extracts ('E-values'). A greenhouse experiment was set up to assess if L/E ratios are affected by P supply and by crop type. The L-and E-values were determined in three P-deficient soils of the Nigerian Northern Guinea savanna (NGS), applied with various rates of TSP, for two cowpea varieties (Vigna unguiculata L., cv Dan-Ila and cv IT-82D-716) and maize (Zea mays L., cv oba super I) as a reference. Plants grown in control soils were severely P-deficient. Plant growth and shoot P uptake significantly increased with increasing P application in all three soils and for all crops, but relative yield and shoot-P responses to P application were similar between maize and cowpea. Both L-and E-values increased with increasing P application. Average L/E ratios for maize were 1.4±0.3 and were unaffected by the P application. For cowpea in contrast, L/E ratios were 3.1±0.2 (significantly larger than one) in one of the three control soils and significantly decreased to 1.3±0.1 at largest P supply. Elevated L/E ratios in cowpea were not associated with an increase in P uptake compared to the other two control soils in which no increase in L/E ratio was observed. It is concluded that cowpea is able to access non-labile P under P-deficient conditions. However, this process cannot overcome P deficiency in these soils, probably because P uptake is limited by the small P concentration in the soil solution (1-2 lg P L )1 ) and this limitation is not overcome by an increase in the accessible soil P quantity (L-value).