Microelectrode measurements of basal, apical and transepithelial potentials in the Malpighian tubules of Drosophila melanogaster were obtained under a range of conditions in order to investigate whether each of the three main second messenger systems known to act in the tubules (cyclic AMP, cyclic GMP and Ca2+) acted specifically on either cation or anion transport, or whether they activated both systems. Ion-selective microelectrode determinations of K+ concentration and pH of secreted fluid allowed the role of each signalling system to be analysed further. Stimulation with cyclic nucleotides markedly alters the potential profile across principal cells through the selective activation of an apical electrogenic V-ATPase. By contrast, manipulation of extracellular chloride levels, combined with stimulation with leucokinin, does not affect the potential profile across the principal cells, showing that chloride must pass through another route. The cell-permeant Ca2+ chelator BAPTA-AM was shown to suppress the action of leucokinins (insect peptides that induce rapid fluid secretion), but not those of cyclic AMP, the neuronally derived insect peptide cardioacceleratory peptide 2b (CAP2b) or its intracellular messenger cyclic GMP. This shows that leucokinins act through Ca2+ and not through cyclic nucleotides and that the cyclic nucleotide pathways do not co-activate the intracellular Ca2+ pathway to exert their effects. Taken together, these results show that leucokinin acts through intracellular Ca2+, independently of cyclic AMP or cyclic GMP, to raise the chloride permeability of the epithelium. By contrast, either cyclic AMP or cyclic GMP (upon CAP2b stimulation) acts on the electrogenic cation-transporting apical V-ATPase, with only a negligible effect on anion conductance and without perturbing intracellular [Ca2+]. There is thus a clear functional separation between the control pathways acting on cation and anion transport in the tubules. Given the evidence from D. melanogaster and other species that chloride does not pass through the principal cells, we speculate that these two pathways may also be physically separated within cell subtypes of the tubules.