Vagal reflexes from the heart and lungs elicit autonomic as well as somatomotor responses. The purpose of the present investigation was to determine whether the inflammatory mediator thromboxane A₂ inhibits the knee-jerk reflex via a vagally mediated reflex from either the heart or the lung. The thromboxane A₂ mimetic U-46619 (0.8 ± 0.08 µg/kg) was injected through a catheter placed near the right atrium (n = 11), near the aortic arch (n = 7), or into the pericardial sac (n = 4) in 11 chloralose-anesthetized cats. The knee-jerk reflex, elicited by striking the patellar tendon with a solenoid-driven hammer, was used to evaluate somatomotor activity. The mean maximum tension produced by the knee-jerk reflex was 306 ± 21 g (range 154-471 g). Intravenous U-46619 injection inhibited the knee-jerk reflex by 25 ± 6% and increased peak systolic pressure 53 ± 7 mmHg on average. Bilateral cervical vagotomy abolished the somatomotor inhibition but did not reduce the pressor response. Intra-arterial U-46619 injection inhibited the knee-jerk reflex in two of seven cats and increased peak systolic pressure by 41 ± 11 mmHg. Vagotomy abolished the inhibition in one of the two cats but did not reduce the pressor response. Intrapericardial U-46619 injection did not affect the knee-jerk reflex nor blood pressure. The results indicate that U-46619 inhibited the knee-jerk reflex via a vagal reflex from the lung because the inhibition predominated after intravenous injection and was abolished by vagotomy. Speculation is made that the inflammatory mediator thromboxane A₂ may contribute via a vagal reflex to the depression of motor activity associated with sickness behavior.