The equivalent dose of body surface can be calculated using a thermoluminescence dosimeter (TLD). The TLD must be calibrated during the initial testing process using the TLD Reader and this process will zero the dose which means the equivalent dose in it will be erased. This study is to read doses for TLD that have not been registered/calibrated in TLD Reader. The sample is 40 calibration cards of TLD Harshaw’s shared in 5 sets testing with different dose, so every sets consists of 8 calibration cards. The calibration card was irradiated with 137Cs source at a dose of 2 mSv, 3 mSv, 5 mSv, 7.5 mSv, and 10 mSv. The calibration curve was created by comparing the irradiation dose with the response calibration readings of the TLD Reader. The best calibration curve is the calibration curve on the polynomial/ quadratic model with R² = 0.9968 its mean pass the reference that good calibration curve have value of R² ≥ 0.990. The standard deviation of the dose measurement results at −3.87% to 21.98%, which means that the value is still in the range of −30% to 50% or not to exceed −30% to 50%. Equivalent dose in uncalibrated TLD can read using a polynomial calibration curve formula as evidenced by the value of R² ≥ 0.990 and a standard deviation that does not exceed −30% to 50%.