Misra et al. (1971) found that 1 h after a single subcutaneous injection of [N-Me-l4C1morphine into the rat, three radioactive compounds could be extracted from the cerebral cortex, namely free morphine, morphine 3-glucuronide and a ‘non-acid-hydrolysable’ conjugate. The regional and subcellular distribution of morphine and its metabolites in brain after subcutaneous injection has never been studied in detail. The work described here was planned to study this distribution and to see if this would provide some information about the central action of opiate drugs. In preliminary experiments rats were injected with [N-Me-14C]morphine hydrochloride (10mg of free base/kg body wt.); 1 h later they were killed and subfractionation of whole brain was carried out to yield the P1, P2, P3 and S3 fractions (nomenclature of Gray & Whittaker, 1962). The amount of morphine, morphine glucuronide and nonhydrolysable metabolites of morphine present in each of the fractions was then assayed. Most of the morphine in the brain was associated with the S3 fraction, and very small amounts of the free drug were associated with the P3 fraction. A similar pattern of distribution of morphine glucuronide and non-hydrolysable metabolites within the subfractions was observed. In order to study the regional distribution of morphine and its metabolites within rat brain, animals which 1 h previously had been injected with radioactive morphine, were killed, their brains were quickly removed and dissected into six discrete regions by using an adaptation of the method of Glowinski & Iversen (1966). The cerebellum was found to contain significantly more morphine/g of brain tissue than did any other region of the brain; the midbrain contained the least amount of free drug. The cerebellum was also found to contain significantly more morphine glucuronide/g of brain tissue than did any other brain region except the medulla, where the increase did not reach statistically significant values. The striatum contained significantly less morphine glucuronide than any of the other brain regions studied. There were no significant differences in the amounts of non-hydrolysable metabolites present in the six brain regions. The uptake of morphine and its metabolites into the brains of morphine-dependent rats after a subcutaneous injection of radioactive morphine has also been studied. Although there was significantly less morphine taken up into the brain regions of the dependent animals when compared with the control rats, the pattern of distribution of morphine between the regions was almost the same in the two groups of animals. Significant decreases in the amounts of morphine glucuronide in the cerebellum, midbrain and cortex of the dependent rats were seen compared with the non-dependent animals. Subcellular-distribution studies in dependent rats showed that after an injection of [14C]morphine there was less morphine and morphine glucuronide associated with any one particular subfraction of the dependent rat than was associated with the corresponding fraction from control animals; the pattern of distribution of morphine and its metabolites in the dependent and non-dependent rats was, however, very similar. In addition to the subcellular and regional studies described above, the subcellular distribution of morphine and morphine metabolites within each of the six brain regions was studied. This subcellular-distribution pattern was similar for all regions, indicating that there was no accumulation of morphine or morphine metabolites in any one subcellular fraction of any one particular brain region (Table 1). The regional distribution of morphine and morphine metabolites described here is especially surprising in view of the evidence presented by Pert & Snyder (1973a,b) and by Teschemacher et al. (1973), which suggests that the cerebellum does not have any functional role in the effects of morphine, but that the medulla, midbrain and striatum (regions shown in our study to be low in morphine) are possibly involved in the actions