The prospective symptom reports of women seeking treatment for premenstrual symptoms and control subjects were investigated. In order to compare symptom reports from premenstrual symptom sufferers and control subjects a method of combining and analysing prospectively collected menstrual cycle symptom data is required. A technique that uses the time of onset of menses and the time of ovulation (as measured by urinary luteinizing hormone excretion) to standardize each cycle into 14 time points was developed. Summary factors were then empirically derived from data collected prospectively from 30 premenstrual symptom sufferers and 19 control subjects. Twenty-two mood symptoms were summarized into a single factor and the 29 most frequently occurring physical symptoms were summarized into two factors. Factor scores were calculated on the basis of these factors and the effect of time during the menstrual cycle on these scores examined. Both physical symptom factor scores increased significantly in the luteal phase for both the premenstrual symptom sufferer group and the control group. The single mood factor score increased significantly in the luteal phase for the premenstrual symptom sufferer group but not for the control group, suggesting that the only qualitative difference between the groups was the presence of cyclic mood symptoms in the premenstrual symptom sufferer group. The premenstrual symptom sufferer group recorded significantly higher scores on each of the three factors than the control group. The correlation between the scores on each of the factors over three cycles was high both in the follicular and luteal phase suggesting that these factor scores provide a reproducible measure of menstrual cycle symptomatology.