Methadone is an opiate drug that has been identified as an in-vitro substrate of the efflux pump P-glycoprotein (P-gp), active in the intestinal epithelium and in the blood–brain barrier (BBB), among other sites. The objective of this study was to test in vivo, in the rat model, the role of P-gp modulation on the analgesic effect and brain uptake of methadone, as well as identify the most relevant site via dual oral and intravenous (i.v.) experiments. The P-gp specific inhibitor (valspodar or PSC833) was preadministered (10 mg kg−1 i.v.) to test groups. Analgesia was measured using the tailflick test. The ED50 for oral methadone (2, 3, 6 and 8 mg kg−1) decreased three-fold in valspodar groups compared with controls (2.23 + 0.002 mg kg−1 and 6.07 + 0.07 mg kg−1; P < 0.0001). The overall analgesic effect (% antinociception) was elevated 3.1 times in pretreated compared with control rats (90.65% + 0.22 vs 29.23% + 14.0; P < 0.01) after 6 mg kg−1 oral methadone and 2.8 times after i.v. (0.35 mg kg−1) administration (91.75% + 4.27 vs 32.45% + 9.0; P < 0.01). The brain:plasma distribution ratio was higher in pretreated animals and AUCbrain (overall brain concentration) was 6 times higher after oral methadone and 4 times higher after i.v. compared with controls, disproportionally increased relative to plasma, implying an active process at the BBB. P-gp, and hence substrate comedication, plays a critical role in the evolution of the methadone analgesic effect and in its brain uptake, independent of the administration route.