In cats anasthetized with chloralose it was found that below about 180 C. the rhythmic pattern of discharge in low-threshold pulmonary stretch fibres is reversed in relation to the respiratory cycle. This reversed pattern is apparently responsible for the reversal of the Hering Breuer inflation reflex (Head's paradoxical reflex). Evidence favouring this is that reflex inhibition of respiration, produced by high frequency electrical stimulation of pulmonary stretch fibres applied against a background of low frequency stimulation, is converted to excitation between 9 and 110 C.Head's paradoxical reflex appears when the total discharge in low-and higherthreshold pulmonary fibres during inflation becomes less than the total discharge in the former before inflation. A nomogram showing the relation between temperature and the peak frequency of a train of impulses in fibres of different conduction velocities is presented. The likely mechanism responsible for the quantitative variation of the inflation reflex (and its reversal) with temperature is given. SOME of our present concepts [cf. Aviado and Schmidt, 1955;Widdicombe, 1964; Wyss, 1964] regarding the reflex effects of cardio-respiratory and cardiovascular afferent fibres have arisen from experiments in which the vagus was cooled to various temperatures [Head, 1889;Hammouda and Wilson, 1935 a; 1935 b;Partridge, 1939; Whitteridge and Biilbring, 1944;Hammouda et al., 1943; Torrance and Whitteridge, 1948;Dawes et al., 1951; 1954 b;1959]. The conclusions in the more recent of these studies (from 1948 onwards) have rested on the tacit or stated assumption that faster conducting fibres are blocked at higher temperatures than the slower fibres. No experimental support for this assumption has been obtained so far. On the other hand, in a recent paper it has been demonstrated conclusively that (apart from individual variations) conduction in all medullated fibres is blocked at about the same temperature of about 80 C. regardless of conduction velocity 1965 b]. Other relevant and important new findings are (1) that the absolute refractory period (ARP) of nerve fibres varies inversely with their conduction velocities, i.e. fibre diameter [HIursh, 1939]; (2) the maximum frequency of a train of impulses that can be conducted in a fibre without any block is determined by the ARP after the second (or subsequent) impulse; and (3) that under appropriate conditions total block of a train of impulses can occur if the frequency of discharge is greater than this maximum transmissible frequency [Paintal, 1965 b].In view of these findings it follows that our concepts regarding the effect of temperature on reflex effects in relation to the sensory input need to be reconsidered. The present paper provides an interesting example for this need. It demonstrates that Head's paradoxical reflex [Head, 1889] can be 151