In recent decades several pasture legumes have been available in southern Australia as potential alternatives to the most widely used annual pasture legume Trifolium subterraneum. Little is known about their soil phosphorus (P) requirements, but controlled environment experiments indicate that at least some may differ in their P fertiliser requirements. In this study, pasture legume varieties, including T. subterraneum as the reference species, were grown at up to four sites in any one year over a 3-year period (in total, seven site × year experiments) to measure herbage growth responses in spring to increased soil P availability. A critical soil test P concentration (corresponding to 95% maximum yield) was estimated for 15 legumes and two pasture grasses. The critical soil P requirements of most of the legumes did not differ consistently from that of T. subterraneum, indicating their soil fertility management should follow the current soil test P guidelines for temperate Australian pastures. However, the critical P requirement of Medicago sativa was higher than that of T. subterraneum, but remains ill-defined because extractable soil P concentrations in these experiments were often not high enough to permit a critical P estimate. Three forage crop legumes (Trifolium incarnatum, Trifolium purpureum, Trifolium vesiculosum) and two pasture legumes (Ornithopus compressus, Ornithopus sativus) had lower critical soil test P concentrations. It may be feasible to manage pastures based on these species to a lower soil test P benchmark without compromising yield.