Using an electrical technique we estimated the thickness of the unstirred layer in the human jejunum during kinetic studies of electrogenic glucose absorption. The unstirred layer in seven healthy volunteers (632 +/- 24 mum: mean +/- SEM) was significantly thicker than in 10 patients with active coeliac disease (442 +/- 23 mum) but not significantly different in seven patients who had responded to treatment by gluten withdrawal (585 +/- 49 mum). There were similar differences in the values of ;Apparent Km' for electrogenic glucose absorption between healthy control subjects (36 +/- 6 mM) active coeliac patients (11 +/- 1 mM) and treated coeliac patients (31 +/- 5 mM). The changes in PDmax however, showed a different pattern. The PDmax in the active coeliac group (6.8 +/- 0.7 mV) was lower than in controls (7.6 +/- 0.6 mV) but not significantly so, while the PDmax in the treated coeliac group (10.6 +/- 0.9 mV) was significantly higher than in both the active coeliac and control groups. It should be noted that both operational kinetic parameters obtained in the present study are much lower than those obtained previously (Read et al., 1976b) because of the use of siphonage. Analysis of the results using a computer simulation indicates that the reduction in Apparent Km in active coeliac disease can be caused by the interaction of the decreased maximal absorption rate for glucose (Jmax) with the attenuated unstirred layer. In these circumstances it is not necessary to postulate any change in the affinity of the transport mechanism for glucose (;Real Km'). It is remarkable that the disease process produces an Apparent Km which is much closer to the Real Km than that found in health.