Therapy with radioactive iodine (131I) is a well established treatment method for postsurgical differentiated thyroid carcinoma (DTC). A fixed discharge time is generally set, regardless of individual differences in residual body radioactivity (RBA). This study aimed to investigate the RBA of each patient to find the attenuation law and to identify underlying factors in order to predict the time point for a safe, scientifically sound discharge plan. A total of 231 DTC patients undergoing 131I treatment were all treated with 3.7 GBq (100 mCi) of 131I. RBA was estimated by measuring the external body dose rate (EDR) at a distance of 1 m from the body surface between 0 and 72 hours after oral administration of 131I. Data from each patient were used to establish a time-EDR value (h-μSv/h) curve. Software was developed to predict the time when a patient's dose equivalent meets the national safety standard by including six time points between 40 and 60 hours. Several factors that might affect that time were analyzed. The EDR attenuation law in patients could be described with a double exponential decay model, and the cutoff value was set as 23.3 μSv/h, upon which the predictive software was developed. Student's t-test showed there was no statistical difference between predicted values and the actual measured values (p > 0.05). Correlation analysis found that serum thyroglobulin, total triiodothyronine, total thyroxine, free triiodothyronine, free thyroxine, thyrotropin, 2- and 24-hour iodine uptake rate of the thyroid, scores of 99mTc-pertechnetate thyroid scan, scores of 131I whole-body scan, scores of ultrasound scan, and gastrointestinal residues were associated with attenuation speed. A further multiple linear regression analysis found that 24-hour iodine uptake (X1), residual thyroid grading by 131I whole-body scan (X2), blood free triiodothyronine (X3) and free thyroxine (X4) predominantly influenced the decline of the EDR. The regression equation was Ŷ = 2.091X1 + 6.370X2 + 4.529X3 + 2.466X4 - 8.614 (F = 44.03, p < 0.01). An effective and convenient method was created to measure and predict the individual safety time for discharge. This could play a significant role not only for scientific hospital discharge planning, rational use of medical resources, and better individualized management, but also in public radiation protection.