The SPAQ is a widely used tool for identifying possible cases of recurrent major depressive disorders with a seasonal pattern. However, its test-retest reliability, sensitivity, specificity, positive predictive value, negative predictive value, efficiency and predictive validity have not previously been formally assessed. Forty-seven subjects who fulfilled the DSMIIIR criteria for a major depressive disorder with a seasonal pattern were traced and re-interviewed after five to eight years. The SPAQ was found to have a positive predictive value of 48% and an efficiency of 57% in identifying cases of SAD confirmed by follow-up. The test-retest reliability was low. There was a mean difference in seasonality score between the first and second test of 3.17 ± 4.7 (mean ± 1 S.D.). Although the SPAQ is a rapid method of collecting information about recent seasonal variation, it has low test-retest reliability and on its own is unable to predict the seasonality of the future course of illness.