Organically managed farm areas in Denmark are expanding and typically contain clover-grass leys that are known to stimulate accumulation of organic matter in arable soils. We compared the C allocation to roots and soil from clover and grass, and determined for how long assimilated C remained mobile in these plant-soil systems. Pots with perennial ryegrass, white clover or a mixture of both were pulse-labelled with 14CO2, and harvested for analyses after 4, 11, 20, and 30 days. 14C losses by shoot respiration stopped within 4 days and after this incubation time the input of assimilated 14C to below-ground compartments was greater in grass (52%) than in clover (36%). During the next 4 weeks, 14C allocation below ground increased in grass (up to 75% at day 30), but remained constant in clover (37% at day 30). In the grass/clover mixture, the below-ground fraction increased to 50% at day 30. In clover, 14C was incorporated sooner into stable plant and soil pools and less was released in rhizodeposition than in grass. This was confirmed by the 14C in the soil microbial biomass that decreased fastest in the clover treatment. Root-derived C compounds of clover probably decomposed faster than those from grass. The larger size and specific activity of the soil microbial biomass in the mixed treatment suggested a stimulating effect of the two plant species on substrate utilisation by the microbial community. This study showed that a 2- to 3-week distribution period is needed before sampling for quantitative estimates of C allocation.