The results are reported of a randomized, double-blind study of two daily oral contraceptives, both containing 3mg norethisterone acetate; one also contained 50 microgram ethinyl estradiol, while the other contained a mixture of micronized 4mg estradiol plus 2mg estriol. The trial was conducted to a common protocol at WHO Collaborating Centres for Clinical Research (CCCR's) in Bangkok, Bombay, Singapore and Szeged. Of the 925 women who entered the study, 458 received the product containing "natural" estrogens and 467 received the product containing ethinyl estradiol. Analysis of admission characteristics showed that the two groups were very closely matched. There was no statistically significant difference in contraceptive efficacy between the two products; both have annual failure rates of approximately 1 per 100 women. The discontinuation rate after one year was 51.5 for the "natural" estrogen product and 48.4 for the one containing synthetic estrogen. Discontinuation for menstrual irregularities was significantly higher for the "natural" estrogen preparation. During both the first three treatment cycles and the tenth to twelfth cycles, the incidence of various menstrual irregularities was significantly higher in the "natural" estrogen contraceptive. Adverse clinical associations were those commonly reported with other combined oral contraceptives for both preparations. It is concluded that the high incidence of menstrual irregularities associated with the particular combination of norethisterone acetate and "natural" estrogens renders it less suitable for general use in family planning programmes than combinations containing synthetic estrogens, such as ethinyl estradiol.