We introduce a non-invasive and accurate method to assess tibialis anterior muscle temperature (Tm) during rest, cycling exercise, and post-exercise recovery using the insulation disk (INDISK) technique. Twenty-six healthy males (23.6 ± 6.2 years; 24.1 ± 3.1 body mass index) were randomly allocated into the 'model' (n = 16) and the 'validation' (n = 10) groups. Participants underwent 20 min supine rest, 20 min cycling exercise at 60% of age-predicted maximum heart rate, and 20 min supine post-exercise recovery. In the model group, Tm (34.55 ± 1.02 °C) was greater than INDISK temperature (Tid; 32.44 ± 1.23 °C; p < 0.001) and skin surface temperature (Tsk; 29.84 ± 1.47 °C; p < 0.001) throughout the experimental protocol. The strongest prediction model (R(2) = 0.646) incorporated Tid and the difference between the current Tid temperature and that recorded four minutes before. No mean difference (p > 0.05) and a strong correlation (r = 0.804; p < 0.001) were observed between Tm and predicted Tm (predTm) in the model group. Cross-validation analyses in the validation group demonstrated no mean difference (p > 0.05), a strong correlation (r = 0.644; p < 0.001), narrow 95% limits of agreement (-0.06 ± 1.51), and low percent coefficient of variation (2.24%) between Tm (34.39 ± 1.00 °C) and predTm (34.45 ± 0.73 °C). We conclude that the novel technique accurately predicts Tm during rest, cycling exercise, and post-exercise recovery, providing a valid and cost-efficient alternative when direct Tm measurement is not feasible.