Labor initiates one of the most intensely painful episodes in a woman's life. Opioids are used to provide analgesia with substantial interindividual variability in efficacy. mu-Opioid receptor (muOR, OPRM1) genetic variants may explain differences in response to opioid analgesia. We hypothesized that OPRM1 304A/G polymorphism influences the median effective dose (ED(50)) of intrathecal fentanyl via combined spinal-epidural for labor analgesia. Nulliparous women were prospectively recruited around 35 weeks gestation (n=224), and genotyped for 304A/G polymorphism. Those requesting neuraxial labor analgesia were enrolled in one of the two double-blinded trials: up-down sequential allocation (SA, n=50) and a separate confirmatory random-dose allocation trial (RA, n=97). Effective analgesia from intrathecal fentanyl was defined by >or=60 min analgesia with verbal rating score <or=1 (scale 0-10) and was compared between mu OR 304A homozygotes (Group A) and women carrying at least one 304G allele (Group G). OPRM1 304G allele frequency f(-) was 0.18. Using SA, intrathecal fentanyl ED(50) was 26.8 microg (95% CI 22.7-30.9) in Group A and 17.7 microg (95% CI 13.4-21.9) in Group G (p<0.001; 304A homozygosity increased the ED(50) 1.5-fold). RA confirmed that 304A homozygosity significantly increases intrathecal fentanyl ED(50) (27.4 microg in Group A and 12.8 microg in Group G [p<0.002; 2.1-fold]). We demonstrate for the first time that the muOR 304G variant significantly reduces intrathecal fentanyl ED(50) for labor analgesia, suggesting women with the G variant may be more responsive to opioids and require less analgesic drugs. These findings for intrathecal fentanyl pharmacogenetics may have implications for patients receiving opioids in other settings.