In many countries back and neck problems are the most common reason for work incapacity. Numerous studies have looked for reliable predictors of return to work (RTW) or not return to work (NRTW) for those incapacitated, but with few exceptions their predictive and detective ability has been limited. We aimed to perform a comparative study of various commonly used health measures ability to predict RTW or NRTW in a cohort of men and women sick-listed for more than 28 days due to low-back pain (LBP) or neck pain (NP). A cohort of 1,575 men and women sick-listed more than 28 days due to back or neck problems was asked to answer ten different health measures (scales) about various aspects of health status (health related quality of life, pain, back function and depressivity) at four occasions during a 2-year period. A statistical diagnostic test was used to study how well the subject's values on these scales would serve as a predictor for work resumption. There was an improvement from day 28 up until 2 years, reflected by the different scales, more pronounced for LBP than for NP patients and for men with LBP compared with women with LBP. For all separately tested scales EuroQol (EQ-5D) had the highest overall ability to predict RTW or NRTW irrespective of gender, diagnoses or duration of the problems. When RTW or NRTW were predicted in a cohort of sick-listed low-back or neck patients, EQ-5D had outstanding properties in this respect irrespective of gender, diagnosis or elapsed time during this 2-year study.