Aims: Mental disorders often go undetected in primary care, for persons awarded disability pension, and in sick-leave certificates. No validity tests of instruments for detection and measurement of mental disorders have been performed in long-term sickness absence (LSA). This is the aim of the present study for Common Mental Disorders — Screening Questionnaire (CMD-SQ). Methods: It is validity tested in a well-defined Danish population comprising all persons on continuous sickness absence just exceeding eight weeks. CMD-SQ is composed of SCL-SOM (somatization), Whiteley-7 (illness worry and conviction), SCL-ANX4 (anxiety), SCL-DEP6 (depression), SCL-8 (emotional disorder), and CAGE (alcohol dependency). Results: Of 2,414 incident persons on long-term sickness absence within one year, 1,121 participated in the study by filling in CMD-SQ and a subsample of 337 was diagnosed by a psychiatric examination including Present State Examination. Psychometric properties such as scalability and monotonicity, construct validity, and predictive validity were analyzed. Further change over time was analyzed. SCL-SOM and Whiteley-7 showed slight psychometric and screening statistics whereas a combined scale of SC-ANX4, SCL-DEP6, and SCL-8, called SCL-8AD, showed the best and excellent psychometric and screening statistics. Conclusions: SCL-8AD, a sub-scale of CMD-SQ, is a promising candidate for screening and measurement of mental disorders in long-term sickness absence.