1. Mast cell populations in rat lung and spleen were characterized by the presence of two specific protease markers, rat mast cell protease I and II, using both histochemical and radioimmunoassay techniques. Three mast cell populations with different size, morphology, and localization were found in lung and spleen and were identified according to the expression of rat mast cell protease I (RMCPI+) or rat mast cell protease II (RMCPII+) or of both proteases (RMCPI/II+). 2. All three mast cell types were in the vicinity of calcitonin-gene-related-peptide-immunoreactive (CGRP+) nerve fibres in controls as well as in rats infected by Nippostrongylus brasiliensis in which a large increase in the number of both RMCPII+ and RMCPI/II+ mast cells was found. Ablation of the CGRP+ fibres by neonatal treatment with capsaicin resulted in a marked increase in the number of RMCPII+ and RMCPI/II+ cells in lung and, even more, in spleen of adult rats. 3. The interaction of mast cells with CGRP+ C-fibres was assessed pharmacologically by evaluation of the effects of histamine H3-receptor ligands known to act on various types of nerve endings, including those of C-fibres. The effects of H3-receptor ligands were assessed in controls, nematode-infected rats and neonatally capsaicinized rats. Mast cell activity was evaluated by measurement of [3H]histamine synthesis from [3H]histidine. In control rats, administration of the H3-receptor agonist (R)-alpha-methylhistamine and antagonist thioperamide, decreased and enhanced respectively [3H]histamine synthesis in lung and spleen, indicating a tonic control of mast cell activity by histamine via H3-receptors. Such effects were not found in the jejunum, although RMCPII+ mast cells are in close apposition with neuropeptide-containing fibres. The effects of the H3-receptor agents were maintained in lung and spleen of nematode-infected rats, but were almost suppressed in capsaicinized rats. 4. It is concluded that the control of mast cells by histamine acting at H3-receptors involves neuropeptide-containing nerves and presumably reflects the operation of a local neuron-mast cell feedback loop controlling processes such as 'neurogenic inflammation'. This loop still functions when mast cells proliferate in an inflammatory condition. These observations suggest that the use of histamine H3-receptor agonists may constitute a novel therapeutic approach to limit excessive inflammatory responses resulting from dysregulation of this feedback loop.