Monitoring parturition is fundamental to avoid mortality of newborn calves. Physiological conditions such as pregnancy and lactation influence the diurnal temperature pattern of dairy cows. However, information on changes in body temperature and their association with calving time in dairy cows is lacking. Therefore, our objective was to evaluate the use of external ruminal and mammary gland temperature as predictors of calving day. Thirty Holstein cows were used and measurements were taken from 15 d before the expected calving day until the actual calving day. The ruminal temperature was recorded approximately 2 h post-feeding after an area of 0.4 × 0.4 m on the left side of the animal (including the paralumbar fossa) was shaved and cleaned. Measurements were made at a distance of 1.0 m using a hand-held portable infrared camera (E4 Box, FLIR Systems, Wilsonville, OR). The mammary gland temperature was assessed around 4 p.m. from the posterior side following the same approach used for rumen. The average temperature was calculated by tracing the shaved area using ThermalCAM Researcher Pro 2.7 software (FLIR Systems). The statistical model included the random effect of block and fixed effect of treatment, time and interactions. Ruminal temperature and DMI were not correlated (P = 0.08; r = 0.08). Ruminal temperature and calving day were negatively correlated (P < 0.01; r = 0.19). Compared with 5 d before parturition, ruminal temperature was 1.8°C (P < 0.01), 1.2°C (P = 0.02), 0.5°C (P = 0.15), 0.2°C (P = 0.22), and 0.0°C (P = 0.32) lower at calving, −1, −2, −3 and −4 d, respectively. Mammary gland temperature and calving day were positively correlated (P = 0.04; r = 0.11). Compared with 15 d before parturition, mammary temperature was 1.4°C (P = 0.02), 1.1°C (P = 0.03), 1.2°C (P = 0.02), 1.0°C (P = 0.05), 1.1°C (P = 0.04) and 0.7°C (P = 0.09) greater at calving, −1, −2, −3, −4 and −10 d, respectively. Overall, dairy cows exhibit a distinct change in ruminal and mammary gland temperature commencing approximately 8 and 1 d before parturition. Thus, monitoring the ruminal temperature externally could be a practical tool to predict calving.