Volunteers given alfentanyl in high doses developed muscle rigidity within a minute of the onset of apnoea, but prior administration of pure O2 ensured that they were not hypoxic (Benthousen et al., 1986). Streisand et al. (1993) found that volunteers avoided rigidity when given fentanyl so long as they remained conscious and could be prompted to breathe. Opiate-induced rigidity might thus be triggered by prolonged lack of respiratory movements rather than by hypoxia. In anaesthetised rats given fentanyl there is little or no rigidity because γ-motoneurones are depressed by the anaesthetic, but γ-motoneurones become excited within a minute of the onset of apnoea, activated by a descending pathway as are γ-motoneurones (Gladden & Breckenridge, 2004). If the rats were ventilated and given 10% O2 in N2, or 10%O2 with 6%CO2 in N2, no response was elicited from γ-motoneurones (Gladden et al., 2001). Further, whenever ventilation was stopped, any small spontaneous respiratory movements inhibited the response, suggesting that, as in humans, lack of respiratory movement rather than hypoxia initiated the response. Surprisingly, however, ventilation with pure N2 gave a comparable response to that elicited by apnoea. Simultaneous, in-continuity recordings were made from S1 or L6 dorsal and ventral roots and a nerve branch to the longissimus caudae muscle in Sprague Dawley rats anaesthetised with urethane (1.7g kg-1 i.p.). They were given fentanyl (50µg kg-1 i.v.) and ventilated. Additional doses of fentanyl were given as required to suppress any spontaneous respiratory movements. Tests of 45s of apnoea were alternated with tests of ventilation with pure N2 for 45s, allowing 5min for recovery between each test. Responses were assessed by rectifying and averaging the multiunit recording from the ventral root. The latencies of responses were the same in both cases (31±2.5s (mean ±S.E.M.) for apnoea; 31±2.4s for N2; n=11 in 4 experiments). The mean magnitudes of the responses were not significantly different (paired t test; mean % increase: apnoea 16±7.8 %; N2 22 ±5.5%). Coincident with this increased traffic in the ventral roots an increased frequency of γ-spikes was recorded in the muscle nerve branch. The same γ-motoneurones appeared to be activated in both conditions as judged by their spike shapes and conduction velocities, and the afferents that they excited. A possible explanation is that a falling pO2 acts as a stimulus initiating the descending drive to the γ-motoneurones.