Aural canal temperature measurement using an ear mould integrated sensor (T(ac)) might be a method suited for continuous non-invasive core temperature estimation in operational settings. We studied the effect of ambient temperature, wind and high intensity exercise on T(ac) and its ability to predict esophageal (T(es)) and rectal temperatures (T(re)). Seven subjects performed a protocol of rest at 21, 10 and 30 °C, followed by exercise and recovery at 30 °C. The subjects performed the protocol twice: with and without face-wind from halfway through the 30 °C rest period. Extra auricle insulation was applied at one side. Ambient temperature changes affected T(ac) significantly, while T(es) and T(re) remained stable. Insulating the auricle reduced but did not abolish this effect. Wind had an immediate cooling effect on T(ac) independent of auricle insulation. During exercise and recovery in 30 °C, T(ac) provided acceptable group predictions of T(re) in trials without wind (bias: -0.66 ± 0.21 °C covered, -1.20 ± 0.15 °C uncovered). Bias was considerably higher with wind, but variability was similar (-1.73 ± 0.11 °C covered, -2.49 ± 0.04 °C uncovered). Individual predictions of T(es) and T(re) showed more variation, especially with wind. We conclude that T(ac) may be used for core temperature assessment of groups in warm and stable conditions.