OBJECTIVES: The aim of the current study was to investigate whether genetic variation in genes across the serotoninergic system is associated with chronic widespread pain (CWP) and the number of pain sites reported. METHODS: A discovery cohort, with pain data at 3 time-points, was used to investigate genetic associations with two phenotypes: 1) CWP (at ≥2 time-points; n=164) compared to pain free controls (at 3 time-points; n=172) and 2) the maximum number of pain sites reported at any one of the 3 time-points (0-29 sites; n=989). A cohort of 2285 men with a DNA sample and pain data (including 203 CWP cases and 929 controls) was used for validation. Pair-wise tagging (r(2) >0.8) SNPs were genotyped. Logistic and zero-inflated negative binomial regression were used to test for SNP associations with CWP and the number of pain sites, respectively. RESULTS: SNPs in HTR2A were associated with both pain phenotypes in the discovery cohort and a number of these were replicated in the validation cohort some of which were attenuated after adjustment for depression. There was an increased odds of having CWP in subjects with one or 2 copies of the T allele of rs12584920 (discovery OR=1.64 (1.01, 2.60) p=0.03 and validation OR=1.46 (1.07, 2.00) p=0.018). A similar association was observed between rs17289394 and the maximum number of pain sites reported in both cohorts. Conducting a meta-analysis of the two cohorts further strengthened these findings. CONCLUSIONS: This study supports the role of HTR2A in the genetic predisposition to musculoskeletal pain.