Study Design. Multiphase cross-sectional survey of musculoskeletal pain.Objectives. To estimate the prevalence of all reported and clinically significant spinal pain. To identify independent predictors of spinal pain.Methods. A total of 5752 adults sampled from three general practice registers were mailed a screening questionnaire. Subjects who reported the spine as a predominant site of pain were sent a site-specific questionnaire (i.e., back or neck) that asked about severity. Prevalence estimates were calculated and extrapolated to the general population. Predictors of spinal pain were identified using logistic regression with comprehensive adjustment for confounders (including pain at other anatomic sites).Results. The 1-month-period prevalence of all reported spinal pain was 29% (95% confidence interval 27-31%), of which about half was intense, half was chronic, 40% was disabling, and 20% was intense, disabling, and chronic. Most people with back (75%) or neck (89%) pain also reported pain at other sites. Age, female gender (neck pain only), high body mass index, living in an area of raised material deprivation, and south Asian ethnicity were significant predictors of spinal pain with disability. The association between body mass index and deprivation and neck pain was lost after adjustment for pain at other sites. However, even after full adjustment, obesity (OR, 1.7; 95% confidence interval, 1.1-2.5) and high deprivation (OR, 1.7; 95% confidence interval, 1.1-2.7) were predictors of back pain with disability.Conclusions. Adjustment for pain at other sites enabled assessment of whether observed associations were with spinal pain itself or with the other sites of pain. Obesity is an important independent predictor of back pain and its severity. This has implications for primary prevention. The prevalence of spinal pain with disability continues to rise into old age. This has implications for healthcare planning.