Two experiments were carried out to investigate phosphorus efficiencies and mycorrhizal responsiveness in an improved cultivar (Clipper) and a landrace (Sahara) of barley (Hordeum vulgare L.). In experiment 1, two pot sizes were used to evaluate the effect of soil volume on P uptake and mycorrhizal responsiveness. In experiment 2, a compartmented ("cross-pot") system was used to monitor (32)P delivery by external hyphae of arbuscular mycorrhizal fungi (AMF) to the host plant. Results showed that, irrespective of growth conditions, Sahara had much larger root biomass than Clipper and consequently substantially more P was allocated to roots in Sahara than in Clipper. Specific root length in Clipper was much longer than in Sahara. Increase in soil volume enhanced percentage root length colonised by AMF, plant growth and P uptake, and Sahara was more sensitive to changes in soil volume than Clipper. Pot size (soil volume) used to assess responsiveness to AMF by different plant species or genotypes with different root/shoot ratios might be a confounding factor. Clipper was more responsive to AMF than Sahara in terms of tissue P concentrations, which is partly related to their differences in root/shoot ratios. However, increases in SPU [specific P uptake, mg P (g root biomass)(-1)] caused by AMF were bigger in Clipper, suggesting that AMF played a larger role in P uptake. In accordance with the larger increase in SPU, Clipper took up more (32)P via AMF hyphae than Sahara. The compartmented system using radioactive P might be an alternative approach to directly investigate mycorrhizal responsiveness of different plant species or varieties than conventional pot experiments, provided that the same AM fungus is used.