A cohort of 243 female sewing machine operators were studied by questionnaire and clinical examinations with the purpose to study the efficiency of a screening questionnaire-based method to identify prevalent and incident cases of neck-shoulder disorders. The cross-sectional correlation between self-reported neck-shoulder complaints scores and clinical signs of a neck-shoulder disorder was high. An assessed cut-of point of complaints could identify 90% of all cases fulfilling the criteria of rotator cuff tendinitis, while the complaint score with regard to clinical signs of myofascial pain syndrome was about 67%. In the follow-up analyses, incident clinical signs of disorders at 1 year follow-up were used as 'golden standard' in the validation of screening criteria, which included a defined increase of symptoms. Apparently increase of symptoms did not seem to be the optimal criteria for identification of incident cases. Alternatively a cut-of point of regional complaints in close relation to clinical examinations is recommended.