The objectives of this study were (1) to determine the net energy requirements for the maintenance of gestating sows based on indirect calorimetry, and (2) to explore the feasibility of predicting the net energy requirements for the maintenance of gestating sows based on daily heart rate monitoring. In Exp. 1, six Landrace × Yorkshire crossbred reproductive sows with an initial body weight of 229.5 ± 14.9 kg at d 56 of gestation were randomly assigned to six diverse energy feeding levels using a 6 × 6 Latin square design. The experimental diet was formulated using corn, soybean meal, and wheat bran as major ingredients, and the six feeding levels were set as 1.2, 1.4, 1.6, 1.8, 2.0, and 2.2 times metabolizable energy for maintenance (100 kcal ME/kg BW0.75·d−1), respectively. The animal trial lasted for six periods with 9 days per period, encompassing 5 days of adaptation, 3 days of calorimetry in fed state, and 1 day of calorimetry in fasting state. In Exp. 2, six Landrace × Yorkshire crossbred pregnant sows with an initial body weight of 232.5 ± 12.5 kg at d 64 were fed a corn–soybean meal diet. All sows were tested in a respiratory calorimetry chamber for a 4 day calorimetry test. The heat production of the gestation sows was measured every 5 min using indirect calorimetry, and the heart rate of the gestating sows was recorded every minute using a belt-shape monitor. The results showed that the net energy requirements for the maintenance of gestating sows significant increased as the gestational stage progressed (p < 0.05), and a linear regression model revealed the average net energy requirement for the maintenance of gestating sows was 410 kJ/BW0.75 d−1 during late gestation (days 70–110). Moreover, the average heart rate of the gestating sows was 84 bpm, and the mathematical model developed to predict the net energy requirements for the maintenance of gestating sows was NEm(kcal/h)=19901+exp[136−HR(bpm)43]. In conclusion, the average net energy requirement for the maintenance of sows during late gestation was 410 kJ/BW0.75 d−1, and the utilization of the heart rate monitoring method was found to provide a relevant, accurate prediction for the net energy requirements of sows.