Several work-related, psychosocial and individual factors have been verified as being related to neck and shoulder pain, but the role of pathology visualized by magnetic resonance imaging (MRI) remains unclear. In this study, the relationship between neck and shoulder pain and cervical high-field MRI findings was investigated in a sample of persons in a longitudinal survey. The study aimed to determine whether subjects with persistent or recurrent neck and shoulder pain were more likely to have abnormal MRI findings of cervical spine than those without neck and shoulder pain. A random sample of 826 high-school students was investigated initially when the students were 17-19 years, and again when they had reached 24-26 years of age. Eighty-seven percent participated in the first survey in 1989, of whom 76% took part in the second survey, in 1996. The validated Nordic Musculoskeletal Questionnaire was used to collect data about neck and shoulder symptoms. Two groups were chosen for the MRI study: the first group ( n=15) consisted of the participants who had reported no neck and shoulder symptoms in either of the inquiries, while the second group ( n=16) comprised those who were suffering from neck and shoulder symptoms once a week or more often at the time of both surveys. The degrees of disc degeneration, anular tear, disc herniation and protrusion were assessed by two radiologists. The differences between the two study groups were evaluated. The study found that abnormal MRI findings were common in both study groups. Disc herniation was the only MRI finding that was significantly associated with neck pain. These findings indicate that pathophysiological changes of cervical spine verified on MRI seem to explain only part of the occurrence of neck and shoulder pain in young adults.