Root growth responses to separately placed of bands of N and P fertiliser were examined at the 3-leaf (GS13) and stem extension growth stages (GS30) for wheat (Triticum aestivum L. cv. Yitpi) growing in 2 major alkaline soil types from the rainfed (375–420 mm) grain production regions of south-eastern Australia. Intact cores of a Sodosol and a Vertosol were destructively sampled and changes in root length density (RLD) and root diameter distribution within the soil profile were examined using restricted maximum likelihood analysis and principal component analysis, respectively. At GS13, RLD increased in the Vertosol when only P was applied, although there was no shoot growth response. The root response to P consisted of a spatially generalised increase in RLD, rather than a specific increase in the vicinity of the P fertiliser band. There was a substantially greater, but still generalised, increase in RLD in the Vertosol when both N and P fertiliser were applied, although there was no response to N fertiliser (without P). The distribution of root length in diameter classes changed with depth in the profile at GS13 but was otherwise similar, regardless of soil types and fertiliser treatment. The root responses to fertiliser at GS30 also consisted of a generalised proliferation of RLD in the topsoil, with no detectable fertiliser-specific changes in the location or structure of the root system. Shoot and root growth increased to a similar level at GS30 when plants were supplied with N, irrespective of P, and root diameter distributions were again insensitive to fertiliser treatment. Plants responded to N by increasing the RLD of relatively fine roots (100–250 μm), which was a P style of acquisition strategy that was possibly triggered by moisture limitations. Consequently, the root responses to fertiliser under realistic semi-arid conditions did not follow expectations based on nutrient acquisition studies. Instead, wheat plants responded to N or P fertiliser with a generalised proliferation of fine roots, apparently to better compete for finite water and nutrients.