Prospective, multicenter, cohort study. To examine which clinical and sociodemographic baseline variables can predict a favorable outcome in subjects with neck pain treated by chiropractors. Relatively little is known on predictors of neck pain, particularly for those subjects undergoing chiropractic care. No previous study has examined predictors of outcome for subjects with neck pain by modeling the trajectories of subjects in a longitudinal design. All new, consecutive patients, between 18 and 65 years of age with neck pain of any duration, who had not undergone chiropractic or manual therapy in the prior 3 months, were recruited. Questionnaires were administered at the first 3 visits, and at 3 and 12 months. In all, 29 putative prognostic baseline variables were evaluated. Multivariate multilevel longitudinal regression analyses were conducted using neck pain, neck disability, and perceived recovery as outcomes. In total, 529 patients fulfilled the inclusion criteria. The response rate at 12-months was 92%. In the multivariate analyses, 14 (48%) of the prognostic variables examined were retained in at least one of the models. Shorter duration of neck pain at the first visit was the only variable retained in all 3 final regression models. The following were predictive of a favorable outcome for any 2 of the 3 outcome measures examined: intermittent neck pain, those not on sick-leave or receiving workers compensation at baseline, a higher level of education, less tiredness, higher expectations that the treatment would be beneficial, lack of morning pain, and worse perceived general health. On the basis of the patient's history, the clinician can identify a number of determinants, which are predictive of a favorable outcome. Shorter duration of neck pain at the first visit was the only variable consistently found to be predictive of a favorable outcome for all 3 outcome measures examined.