1 The bronchoconstriction caused in the guinea-pig by arachidonic acid (AA), bradykinin, adenosine diphosphate (ADP) and adenosine triphosphate (ATP) was correlated with effects on platelets. ATP and ADP produced a brief thrombocytopenia and AA a more prolonged one. Bradykinin had no effect on platelets.2 Aspirin inhibited bronchoconstriction and thrombocytopenia produced by AA and part of the bronchoconstriction produced by ATP, but had no effect against ADP. Thrombocytopenia produced by ADP and ATP was not affected by aspirin or indomethacin.3 Platelet depletion by antiserum prevented bronchoconstriction in response to ADP and to ATP, but not in response to bradykinin or to AA, showing that platelets are not involved in aspirin-sensitive bronchoconstriction. Infusions of ADP reduced bronchoconstriction and thrombocytopenia in response to ADP itself and to ATP, but not to AA. Bronchoconstriction by ADP or ATP involves an action on platelets. Only that due to ATP is partially dependent on the activity of prostaglandin synthetase.4 ATP induced aggregation in vitro in guinea-pig platelet-rich plasma (PRP). Rabbit PRP responded only when ATP was first incubated with guinea-pig plasma. The aggregating compound formed was probably ADP, since it was destroyed by apyrase. Its formation was not inhibited by aspirin or indomethacin, indicating that aspirin inhibits ATP-induced bronchoconstriction by a different mechanism.5 The aggregating effect of ATP on guinea-pig platelets was inhibited by concentrations of apyrase that block ADP-induced aggregation, and potentiated by lower concentrations of apyrase.6 Adenosine 5'-tetraphosphate did not aggregate platelets in vivo or in vitro. In vitro aggregation occurred when apyrase was added, suggesting transformation into ADP. Adenosine 5'-tetraphosphate and apyrase inhibited aggregation due to ADP, but failed to affect that due to AA. This suggests that aggregation involving products of prostaglandin synthesis does not require ADP.7 Salicylic acid did not interfere with bronchoconstriction or aggregation due to AA, but prevented inhibition by aspirin when the weight ratio, salicylic acid:aspirin was 4:1. Salicyclic acid may be useful in studies of potential inhibitors of thromboxane A2 synthesis and of thromboxane A2-dependent processes in vivo and in vitro.