Background Psychiatric symptoms are frequent in the perimenopause. They are similar to symptoms observed at different stages of a woman's life cycle, suggesting that there may be an association between mental disturbances of the perimenopause and those observed during premenstrual and postnatal periods. Method This study aimed to determine the reliability of using a modified version of the Steiner premenstrual tension self-rating scale (PMTS) questionnaire for assessing retrospectively the presence of premenstrual complaints and to evaluate the association between previous premenstrual complaints and psychiatric symptoms at the time of the menopause. Forty-one perimenopausal women were selected to establish the reliability of the questionnaire to assess premenstrual symptoms retrospectively (4-8-week interval between measures); agreement between measurements was assessed using the K statistic. Ninety-six women were later recruited from a gynecological menopause outpatient clinic to study the association between premenstrual complaints and the presence of psychiatric symptoms at the time of the menopause (as measured by another self-reporting questionnaire, the SRQ-20); SRQ-20 total scores greater than 7 were considered to be indicative of psychiatric morbidity. Results All 36 PMTS items showed moderate to very good test-retest reliability (0.44 < κ < 1.0). There was a significant correlation between total PMTS and SRQ-20 scores (Spearman correlation coefficient 0.75, p < 0.001), correlation coefficient and SRQ-20 total scores greater than 7 were found in 47.9% of patients. Conclusions Premenstrual symptoms can be reliably measured in perimenopausal women. Women who report having experienced premenstrual dysphoria are more likely to present with psychiatric symptoms at the time of the menopause.