Strong opioids, morphine and oxycodone, are used in the management of moderate to severe cancer pain, but there is wide inter-individual variation in patients' response to these drugs. Here we identify the clinical and pharmacogenetic factors influencing individuals' response and build a model that predicts a favourable clinical outcome to these pain relieving strategies. Two clinical studies were conducted. Firstly, 298 subjects were recruited to a prospective, observational case-control study; data were obtained on 205 morphine responders and 93 morphine non-responders, also known as "switchers‟. Secondly, a randomised control trial was established to compare the clinical response rates to morphine and oxycodone used as first line strong opioids for cancer pain; analysis was carried out on 100 subjects randomised to receive trial medications "A‟ and "B‟ respectively. Clinical, haematological, biochemical, metabolite and genotype data were collected on all subjects; metabolite samples were processed using high performance liquid chromatography and genotyping of single nucleotide polymorphisms (SNP) in candidate genes was performed using sequence specific primer assays whilst other data were obtained by genome-wide SNP analysis. Clinical factors associated with response to strong opioids included age, plasma morphine and metabolite concentrations, concomitant medications and the presence of visceral pain. Genetic variation in the gene encoding the glutamate receptor subunit 2A was associated with response to morphine and a number of polymorphisms were associated with secondary clinical phenotypes. There was no difference in the number of subjects responding to trial medications but dose conversion ratios between trial medications were asymmetric suggestive of cross-tolerance. These findings support a clinical and pharmacogenetic basis of response to strong opioids but the factors influencing response to pain relieving strategies need further definition.