When one ovary of a mouse is removed, the number of eggs shed by the female does not diminish, since the ovulation rate of the remaining ovary doubles. The increased stimulus to the single ovary might come (1) from compensatory growth of the organ enabling it to take up more follicle-stimulating hormone (FSH) from the blood in a given time, or (2) from a compensatory increase in FSH output from the pituitary, or (3) from consumption by the remaining ovary of an increased proportion of the circulating FSH. The experiments described below support the third of these three alternatives. A two-fold increase in ovulation rate by the remaining ovary was already apparent three days after the removal of its partner. Compensatory growth of the ovary, on the other hand, occurred relatively slowly, and a two-fold increase in weight was not achieved for several months after operation. For a given ovarian weight, the single ovary shed more eggs than before the removal of its partner. Compensatory growth cannot therefore be invoked to account for ovulatory compensation; it is more likely to be the result of ovulatory compensation. Five days after the removal of both ovaries, the gonadotrophin content of the pituitary was estimated by means of a mouse uterus assay, and was found to be increased significantly above the control level. By contrast, no increase was detected either 5 or 21 days after the removal of a single ovary. This suggests that the increased number of eggs shed by an ovary after its fellow has been removed is not due to increased FSH output from the pituitary. An attempt was made to see whether transplantation of one ovary to the spleen would lead to an increase in ovulation rate of its partner, suggesting involvement of the pituitary. The experiment failed, since the transplanted ovaries atrophied, whether inserted into the spleen or, in a control group, into the kidney. Ovulation in control and 1-ovary females was induced by gonadotrophin treatment. The response in terms of number of eggs shed per ovary was obtained for control females over a wide dose range of pregnant mare's serum (PMS), and compared over a more restricted range with the response of 1-ovary females. At low doses the 1-ovary females responded at nearly twice the control level, as expected on the hypothesis that ovulatory compensation is due to increased FSH consumption by the single ovary; but the dose-response line was less steep in the 1-ovary than in the control females, so that at high PMS doses the single ovary shed no more eggs than did each ovary of a control female. When the PMS dose was split into two equal parts, the effect of the second part decreased the later it was given. When the dose was split into six equal parts given over a 24-hour period, the mean number of eggs shed showed little or no decline, but the variation in response between animals dropped sharply. One-ovary females treated with 2 i.u. of PMS in a single dose shed fewer eggs than did control females. But when the same dose was given in six equal parts, the response of the 1-ovary animals significantly increased, reaching the control level and thus simulating the physiological situation in respect of ovulatory compensation. This suggests that FSH is being consumed by the ovary in the process of ovulation, and that it is the duration of the hormonal stimulus that determines the number of eggs shed, rather than the absolute level of hormone in the blood.