This study was conducted to determine the relationship between rectal temperature (TREC) and rumen temperature (TRUM) and to assess if TRUM could be used as a proxy measure of core body temperature (TCORE) in feedlot cattle. Eighty Angus steers (388.8 ± 2.1 kg) were orally administered with rumen temperature boluses. Rumen temperatures were recorded at 10-min intervals over 128 days from all 80 steers. To define the suitability of TRUM as an estimation of TCORE, TREC were obtained from all steers at 7-day intervals (n = 16). Eight feedlot pens were used where there were 10 steers per pen (162 m2). Shade was available in each pen (1.8 m2/animal; 90% solar block). Climatic data were recorded at 30-min intervals, including ambient temperature (TA; °C); relative humidity (RH; %); wind speed (WS; m/s) and direction; solar radiation (SR; W/m2); and black globe temperature (BGT; °C). Rainfall (mm) was recorded daily at 0900 h. From these data, temperature humidity index (THI), heat load index (HLI) and accumulated heat load (AHL) were calculated. Individual 10-min TRUM data were converted to an individual hourly average. Pooled mean hourly TRUM data from the 128-day data were used to establish the diurnal rhythm of TRUM where the mean minimum (39.19 ± 0.01 °C) and mean maximum (40.04 ± 0.01 °C) were observed at 0800 h and 2000 h respectively. A partial correlation coefficient indicated that there were moderate to strong relationships between TRUM and TREC using both real-time (r = 0.55; P < 0.001) and hourly mean (r = 0.51; P < 0.001) TRUM data. The mean difference between TREC and TRUM was small using both real-time (0.16 ± 0.02 °C) and hourly mean TRUM (0.13 ± 0.02 °C) data. Data from this study supports the hypothesis that TRUM can be used as an estimate of TCORE, suggesting that TRUM can be used to measure and quantify heat load in feedlot cattle.