A standardized examination protocol was developed for the assessment of neck and shoulder disorders in primary health care. We investigated to what extent the symptoms and signs can predict sick leave due to neck and shoulder disorders. Our study involved 474 patients seeking medical advice from an occupational health service. Most of the 20 symptoms and 15 signs of the protocol were associated with the number of sick leave days within the 60-day postexamination period, but only three symptoms (short duration, high pain intensity, and continuous pain) and two signs (pain in the upper limb during rotation of head and pain in the shoulder during abduction of arm) retained their significance in a multivariate model. No unambiguous subset of protocol items can be suggested for general use. The predictive validity should be tested in other populations and settings before a conclusion is drawn about the external validity of the results.