Abstract An experimental system has been developed for studying efflux of 14C assimilates in growing potato tubers. Small wells are cut into the phloem-rich perimedulla and filled with trap solutions of varying composition which inhibit or promote assimilate efflux. One well on each tuber acts as the treatment while a second well acts as the control. Movement of 14C into wells occurred at comparable rates to that found in intact tissue, harvested from importing tubers in the form of microcores. Sucrose was the predominant translocated sugar in the stolon and was not hydrolysed in either the wells or the microcores following unloading. Efflux into wells containing agar traps was stimulated 40-fold relative to buffer controls by the addition of 20 mol m−3 EGTA to the agar. This was interpreted as passive efflux to the apoplast due to increased membrane permeability in the pathway between the sieve elements and the collecting wells. The EGTA stimulation was reversed by addition of Ca2+. 14C efflux into buffered solutions was inhibited significantly by both DNP and PCMBS, suggesting the involvement of active and carrier-mediated transport components. However, it was not possible to determine whether these compounds acted at the site of unloading only, or on the short-distance transfer step between phloem and collecting wells. The rate of tracer efflux was not significantly different when 1 mol m−3 and 300 mol m−3 sucrose were applied to the wells, indicating insensitivity of solute movement to low apoplastic solute concentrations. However, raising the solute concentration to 800 mol m−3 caused a severe inhibition of tracer efflux. These results were duplicated with mannitol as the osmoticum. It is suggested that plasmolysis prevented further efflux by disruption of a predominantly symplastic transport pathway between the phloem and collecting wells.