Plant growth enhancing effects of arbuscular mycorrhizal (AM) fungi are suitably quantified by comparisons of mycorrhizal and non-mycorrhizal plant growth responses to added phosphorus (P). The ratio between the amounts of added P required for the same yield of mycorrhizal and non-mycorrhizal plants is termed the relative effectiveness of the mycorrhiza. Variation in this relative effectiveness was examined for subterranean clover grown on a high P-fixing soil. Plants were either left non-mycorrhizal or inoculated with one of three AM fungal species with well-characterised differences in external hyphal spread. With no P added, plants from all treatments produced < 10% of their maximum growth achieved at non-limiting P supply. The growth response of non-mycorrhizal plants was markedly sigmoid. Mycorrhizal growth responses were not sigmoid but their shape was two-phased. The first phase was an asymptotic approach to 25-30% of maximum growth, followed by a second asymptotic rise to maximum growth. Growth effects of Glomus invermaium and Acaulospora laevis were quite similar. Plants in these treatments produced up to four times greater shoot dry biomass than non-mycorrhizal plants. Scutellospora calospora was less effective. The relative effectiveness of AM fungi varied with the level of P application. This is expected to apply to all soils on which a sigmoid response is obtained for growth of non-mycorrhizal plants. In a simple approximation the relative effectiveness was calculated to range from 1.46 to 15.57. Shoot P contents were increased by up to 25 times by A. laevis, significantly more than by the other two fungi. The further mycelial spread of this fungus is thought to have contributed to its relatively greater effect on plant P content.