Experiments were performed upon twenty-eight male Californian rabbits anaesthetized with sodium pentobarbitone and breathing room air. Successive injections of glass bead emboli (diameter 125 microns) caused an increase in respiratory rate, and decreases in tidal volume (tachypnoeic response) and platelet count. The tachypnoeic response was not observed in rabbits that had been rendered thrombocytopaenic by platelet antiserum. Induction of thrombocytopaenia by antiserum was accompanied by tachypnoea which was not significantly different (P greater than 0.05) from the response to emboli. The tachypnoeic response to antiserum, like that to emboli, was prevented by bilateral cervical vagotomy. Pre-treatment with aspirin (5 mg/kg), which inhibits the platelet-release reaction, also prevented the tachypnoeic response to both emboli and antiserum, although it did not prevent the fall in platelet count which following injection of antiserum. These data support the hypothesis that the post-embolic reflex tachypnoea is platelet dependent and may be secondary to the platelet-release reaction.