Phosphor-us deficiency is a very common problem in the acid soil of central China. Previous research has shown that starter N and N topdressing at the flowering stage (RI) increased soybean (Glycine max) yield and N 2 fixation (Gan et al., 1997, 2000). However, there is little information available concerning soybean response to P-fertiliser in soybean production in central China (Gan, 1999). A field experiment was conducted to investigate the response to P (0 kg P ha(-1), 22 kg P ha(-1), 44 kg P ha(-1) before sowing) and N fertiliser application (NI: 0 kg N ha-1, N2: 25 kg N ha(-1) before sowing, N3: N2 + 50 kg N ha(-1) at the V2 stage and N4: N2 + 50 kg N ha(-1) at the RI stage) on growth, yield and N 2 fixation of soybean. Both N and P fertiliser increased growth and seed yield of soybean (P < 0.01). Application of basal P fertiliser at 22 kg P ha(-1) or 44 kg P ha(-1) increased total N accumulation by 11% and 10% (P < 0.01) and seed yield by 12% and 13% (P < 0.01), respectively, compared to the zero P treatment. Although application of starter N at 25 kg N ha(-1) had no positive effect on seed yield at any P level (P > 0.05), an application of a topdressing of 50 kg N ha(-1) at the V2 or RI stage increased total N accumulation by 11% and 14% (P < 0.01) and seed yield by 16% and 21% (P < 0.01), respectively, compared to the zero N treatment. Soybean plants were grown on sterilised Perlite in the greenhouse experiment to study the physiological response to different concentrations of phosphate (P 1: 0 mm; P2: 0.05 mm; P3: 0.5 mm; P4: 1.0 mN) and nitrate (N1: 0 mm with inoculation, N2: 20 mm with inoculation). The result confirmed that N and P nutrients both had positive effects on growth, nodulation and yield (P < 0.01). The relative importance of growth parameters that contributed to the larger biomass with N and P fertilisation was in decreasing order: (1) total leaf area, (11) individual leaf area, (iii) shoot/root ratio, (iv) leaf area ratio and (v) specific leaf area. The yield increase at N and P supply was mainly associated with more seeds and a larger pod number per plant, which confirmed the result from the field experiment.