Moulting fluid of Manduca sexta contains high concentrations of potassium and bicarbonate (100 mM) and low concentrations of chloride (5 mM). This fluid begins to disappear from the exuvial space approximately 9-10 h before the actual shedding of the integument. During this time, the integument can be isolated in an Ussing cell and electrical properties measured in vitro. In a normal 32 mM KHCO3 saline, potential difference (PD) is around 10 mV, exuvial side positive, and short-circuit current (SCC) is 15-20 microA cm-2. Substitution of chloride slightly reduces both PD and SCC, although resistance does not change significantly. Measurement of chloride transport in the absence of K+ indicates that 100% of the SCC can be accounted for by the net chloride flux (approximately 2 microA cm-2). The Km and Jmax for transepithelial chloride transport are 14 mM and 0.1 microEq cm-2 h-1. Bilateral potassium addition stimulates chloride transport, doubling net chloride flux as potassium concentration increases from 2 to 5 mM. Chloride net flux is not inhibited by the presence of furosemide (1 mM), nor in HCO3- -free saline by thiocyanate (1 or 10 mM) or acetazolamide (0.1 mM), but is inhibited by 100% N2. The pattern of chloride transport in M. sexta is similar to that previously reported for the rectum of locusts. As chloride is normally at low concentrations in the moulting fluid, it is suggested that this transport system acts to maintain low intracellular concentrations which may be necessary for enzymatic functions in the epidermal cells and has little importance in fluid transport.