The purpose of this study was to determine whether core temperature (T(c)) telemetry could be used in ambulatory women to track changes in the circadian T(c) rhythm during different phases of the menstrual cycle and, more specifically, to detect impending ovulation. T(c) was measured in four women who ingested a series of disposable temperature sensors. Data were collected each minute for 2-7 days and analyzed in 36-h segments by automated cosinor analysis to determine the mesor (mean temperature), amplitude, period, acrophase (time of peak temperature), and predicted circadian minimum core temperature (T(c-min)) for each cycle. The T(c) mesor was higher (P < or = 0.001) in the luteal (L) phase (37.39 +/-0.13 degrees C) and lower in the preovulatory (P) phase (36.91 +/-0.11 degrees C) compared with the follicular (F) phase (37.08 +/-0.13 degrees C). The predicted T(c-min) was also greater in L (37.06 +/- 0.14 degrees C) than in menses (M; 36.69 +/- 0.13 degrees C), F (36. 6 +/- 0.16 degrees C), and P (36.38 +/- 0.08 degrees C) (P < or = 0. 0001). During P, the predicted T(c-min) was significantly decreased compared with M and F (P < or = 0.0001). The amplitude of the T(c) rhythm was significantly reduced in L compared with all other phases (P < or = 0.005). Neither the period nor acrophase was affected by menstrual cycle phase in ambulatory subjects. The use of an ingestible temperature sensor in conjunction with fast and accurate cosinor analysis provides a noninvasive method to mark menstrual phases, including the critical preovulatory period.