This study determined the disposition of morphine in plasma and cerebrospinal fluid (CSF) and its relationship to ventilatory depression. 14C-Morphine sulfate (0.3 mg/kg or 2.0 mg/kg as the base) was injected intravenously into dogs anesthetized with enflurane-N2O-O2. Arterial plasma, cisternal CSF, and urine were analyzed for unchanged morphine and morphine metabolites. By 10 min after injection, 97 per cent of the injected morphine had been cleared from plasma. The terminal elimination half-times for morphine in plasma were not significantly different for the two doses and averaged 75 ± 5 (SEM) min. The volumes of distribution were significantly greater for 0.3 mg/kg than for 2.0 mg/kg morphine; this difference appeared to be related to the lesser hypotension following the smaller dose. Metabolism of morphine was rapid and efficient; metabolites were apparent in plasma by 1.5 min. Morphine conjugates, primarily morphine glucuronide, accounted for more than 90 per cent of the radioactivity in plasma at two hours. By six hours, 66 ± 3 per cent of the administered 14C radioactivity had been recovered in the urine, but only 13 ± 2 per cent had been excreted as unchanged morphine. Morphine concentrations in CSF did not peak until 15–30 min after injection, and the terminal half-time was significantly greater than that of morphine in plasma. Ventilatory depression was prolonged, and its abatement did not correlate with the decline of morphine concentrations in either plasma or CSF. It is concluded that morphine was rapidly cleared from the plasma by its uptake into tissues and by its efficient conjugation with glucuronic acid. Ultimate elimination of the drug was primarily by renal excretion of conjugated morphine. Uptake of morphine into CSF was delayed and its elimination from CSF was prolonged in comparison with its clearance from plasma. The ventilatory depressant effects of morphine were not directly related to its concentration in either CSF or plasma.