The immediate and late asthmatic reactions provoked by inhaled allergens have provided useful models enabling the dissection of individual inflammatory cells and their mediators that may contribute to the pathogenesis of asthma. The immediate reaction is considered to be mast cell-mediated on the basis that about 50% of the response is inhibitable by potent and selective H1-receptor antagonists such as terfenadine and astemizole. Additional inhibition (∼30%) by the potent cyclooxygenase inhibitor flurbiprofen implies an important role for prostanoids in the immediate response, the most likely mast cell-derived product being prostaglandin (PG) D2. In man, PGD2 is selectively metabolised to 9α1 1β-PGF2, a unique prostaglandin which shares with PGD2 contractile properties on guinea-pig and human airways smooth muscle. The inability of piriprost, a potent leukotriene synthesis inhibitor, to influence the allergen-provoked immediate reaction raises the possibility that sulphidopeptide leukotrienes play a minor role in this response. The late asthmatic reaction is considered to resemble clinical asthma since it is accompanied by increased responsiveness of the airways to a wide range of stimuli. The late reaction in man is inhibited by nedocromil sodium (4 mg) but only marginally attenuated by salbutamol (200 μg) if both drugs are administered prior to allergen challenge. Since salbutamol, in the dose administered, is a potent mast cell-stabilising agent, these findings must question the obligatory role of mast cell mediator release in the pathogenesis of the late response. A model of the late reaction in conscious guinea-pigs demonstrates a selective influx of neutrophils followed by eosinophils 6–12 h after challenge in relation to the late reaction, suggesting a role for these secondary effector cells in mediating this response and possibly the acquisition of increased bronchial responsiveness. These human and animal studies demonstrate a complex interaction of primary and secondary effector cells in the pathogenesis of bronchoconstrictor responses to allergen and should help to clarify the contribution of ‘inflammation’ to clinical asthma.