A sample of 80 patients at a mental health institute were rated on a modified NMZ Scale. The medical factor and age factor were changed into more appropriate forms for such a population. The educational and occupational factors were left in the original NMZ Scale. The sample was followed up, at least 12 months later, to determine the patients' vocational status. Using a multiple correlation with this scale, an r of .75 was determined in predicting later vocational status, with only two items remaining out of those used. A cross‐validation check using 30 patients incorrectly identified six people in terms of their later vocational status and produced a significant phi coefficient of .53. The results are discusse