Writing as one with a long-standing interest in the process of mammalian ovulation,1, 2 any new evidence as to factors underlying the release of an oocyte from a mature Graafian follicle is welcome. We now wish to report that one such factor is temperature of the antral fluid, and that this temperature corresponds closely with, or mirrors, that of the ovarian follicular surface. There is a cumulative body of evidence that pre-ovulatory follicles may be more than 1.0°C cooler than ovarian stroma in eutherian mammals, and both compartments are cooler than jugular and deep rectal temperatures.3, 4 Greater confidence in the work and its potential significance has now been achieved in mature cows.5-7 Our recent findings can be summarized in a single sentence. Follicles having a reduced antral fluid temperature compared with that of neighboring tissues invariably ovulated whereas follicles not revealing such cooling did not. This evidence comes from cow ovaries which, it should be noted, have various features in common with human ovaries: these include the pre-ovulatory follicular diameter (commonly 14-22 mm) and the time interval from the gonadotropin peak to ovulation itself (approximately 26-36 hours). Previous studies in rabbits (n = 40),8 pigs (n = 89),1, 9 cows (n = 6)10 and women (n = 26)11 employed infra-red sensing of surface temperatures8-10, 12 or follicular puncture with thermoelectrodes11 in ovaries exposed during mid-ventral or flank laparotomy under systemic or local anesthesia. All follicles in the different studies except one (from an infertile woman)11 were cooler than the ovarian stroma. In a series of recent studies, by contrast, a fine thermistor probe was introduced into the follicular antrum of dairy cows (n = 80) without the use of sedatives, anesthetics or surgery.5-7 The only tissue invasion involved passing a 17G needle containing the probe across the anterior vaginal wall followed by puncture of the ovarian surface. The procedure was monitored under ultrasonic scanning. None of the animals revealed any form of discomfort or stress during or after this straightforward approach. Results are summarized in Table 1. Graafian follicles that ovulated within 24 hours of temperature measurement were a mean of 1.12°C cooler than deep rectal temperatures, whereas no significant differences were found for non-ovulating follicles. This lower temperature was a consistent observation throughout the studies,5-7 enabling follicular cooling to be correlated positively with the potential for pregnancy.7 Evidence in support of this point can be detected in infra-red images of pig ovaries.12 By contrast, the temperature of newly ovulated follicles—ie, without their content of cumulus-oocyte complex and follicular fluid—was similar to that of other ovarian tissues. Pregnancies n (%) The specific mechanisms involved in follicular cooling have not been fully revealed, although they currently focus on endothermic reactions within pre-ovulatory follicles and would require effective local heat exchange.13, 14 The present results should stimulate further examination of antral fluid contents, especially large glycoprotein molecules, during the remodeling that leads to increased viscosity. At a more practical level, these results provide much-needed inspiration for human in vitro fertilization clinics because they offer a meaningful route to oocyte selection in protocols of multiple follicle stimulation by systemic gonadotropin treatment.15 Of course, non-invasive infra-red scanning of ovarian surface temperature would be a more attractive approach than multiple follicle puncture, but this awaits an endoscope capable of transmitting infra-red radiation. Relationships between temperature, follicular growth in vitro and oocyte quality would also be considered in the procedures of fertility preservation.16-19