Mycelial cord systems, up to 50 cm in diameter, of the basidiomycete Phanerochaete velutina, a common woodland saprotroph, were grown on unsterile soil in model laboratory microcosms from either 4 or 16 cm3 wood inocula. After 48 d, systems were supplied at 10-d intervals with pairs of new 4 cm3 beechwood resources placed behind the foraging colony margin, where possible on opposite sides of the system. Image analysis was used to quantify radial extension, hyphal cover, the mass (DBM) and surface fractal (DBS) dimensions of the mycelial systems and wood-resource bleaching activity. Mycelial systems developing from small inocula had significantly (P[les ]0.05) lower radial extension rates, hyphal cover and DBM[ratio ]DBS ratio than those from large inocula. Initially, systems developing from small inocula also displayed significantly (P<0.05) slower wood-resource bleaching activity than those from large inocula, suggesting that carbon limitation affected both foraging behaviour and resource utilization. A separate central compartment containing the inoculum was supplied with 32P orthophosphate and its partitioning amongst wood resources was monitored nondestructively for 44 d. Total 32P acquisition by wood resources was not significantly (P>0.05) affected by inoculum resource size. However, the proportion of total acquired P allocated to resources, which varied according to the length of time that resources had been in contact with the mycelium, was dependent upon inoculum size. The results support the hypothesis that phosphorus translocation is not a demand-driven process. We suggest that the time taken before the greatest rate of phosphorus withdrawal from a uniform translocation stream is dependent upon prior availability of carbon within colonized resources.