Abstract The extracellular space of rat superior cervical ganglia in vitro was measured using mannitol and sucrose labelled with tritium and carbon-14. The volumes of distribution of the 3H-labelled derivatives, especially [ 3H]mannitol, exceeded those of the 14C-derivatives. The divergence increased with increasing lengths of incubation. Thus, after 30 min incubation, ‘spaces’ (ml · g −1) were: [ 14C]mannitol, 0.407; [ 3H]mannitol, 0.447; [ 14C]sucrose, 0.396; [ 3H]sucrose, 0.439. After 120 min, corresponding values were: [ 14C]mannitol, 0.458; [ 3H]mannitol, 0.645; [ 14C]sucrose, 0.430; [ 3H]sucrose, 0.497. Using thin layer chromatography, it was shown that an average of 22% of the label in ganglia incubated for 120 min with [ 3H]mannitol, but only 4% with [ 14C]mannitol, was not associated with the parent compound. Both [ 3H]- and [ 14C]sucrose appeared to be metabolized by 11%. It is concluded that mannitol and sucrose can be metabolized in isolated ganglia and that this may lead to substantial errors in estimating the extacellular space, particularly when [ 3H]markers are used.