Current methodologies to predict methane (CH4) production have disadvantages such as high acquisition and implementation costs, the impossibility of reconciling animal evaluation at pasture with more accurate results, and time needed for measurements. Infrared thermography may be an alternative technology for measuring the amount of CH4 produced and has the advantages of low cost, speed and efficiency in obtaining results. The objective of the study was to determine if the infrared thermography is adequate for predicting the emission of CH4 in hair sheep and the best time after feeding to carry out the measurement. Twelve Santa Inês lambs remained for two days in respirometric chambers, in a semi-closed system, to determine the CH4 production. The animals were divided into two treatments, according to the diet provided. During this period, seven thermographic photographs were taken, at times − 1h, -0.5h, 0h, 0.5h, 1h, 2h, 3h, 5h, and 7h, according to the feeding time, defined as 0h. CH4 production was measured over 24 hours. Thermographic images were used to measure the maximum, minimum, average and point temperatures at the left and right flanks. At each time, the temperature difference between the left and right flanks (left minus right) was calculated. Pearson correlation coefficients, multiple regression and principal component analysis, were carried out in SAS®. For sheep, the time three hours after feeding was considered ideal to predict the CH4 production with the use of infrared thermography images. At hour seven post-feeding, a significant increase in temperature differences were observed, suggesting a new peak of CH4 production.