Several opioids such as codeine are considered pro-drugs as they are metabolized by the highly polymorphic CYP2D6 to more potent opioids with clinical efficacy and toxicological sequelae. Many opioids are also metabolised by CYP3A4/5. In over 600 cancer patients using fentanyl patches, plasma fentanyl concentrations and the metabolic ratio (norfentanyl/fentanyl) were associated with CYP3A4*22 and CYP3A5*3 genotypes but their overall contribution to variability was < 1%. Pharmacogenomic studies on pharmacodynamic variants remain a poorly investigated field; for opioids, the OPRM1 reduced function A118G variant seems not to play a major role in contributing to the variability in opioid dosage in pain patients, possibly due to low frequency (<15%) and low patient numbers in most studies. In over 950 Chinese, Malay and Indian women undergoing caesarean section in Singapore, in whom the A118G variant frequency is over 40%, morphine PCA requirements were significantly related to the G variant allele, highlighting the need to include multiethnic cohorts. As opioids also cause neuroinflammation via innate immune-mediated mechanisms, in both of the above studies (cancer-fentanyl and surgery-morphine), opioid requirements were associated with variants in TLR4 and some of the proinflammatory cytokine gene variants, in an ethnicity-dependent manner. To potentially translate the science into the clinic, we need to look beyond dispositional pharmacogenomics and incorporate drug target and intracellular signaling pathway genetics into our studies and appreciate that including multiple ethnic cohorts may increase our understanding of factors contributing to variability in opioid efficacy.