In current NRC regulations, three options exist that may be used to determine release criteria for patients administered radioactive materials. Absorbed dose estimates may be based on administered activity, measured dose rate, or on patient-specific calculations. All of these methods proposed by the NRC can lead to overestimation of the dose equivalent to others due to their oversimplified nature. The primary oversimplifications are the use of a point source methodology and using the measured surface entrance dose rate to determine whole body dose. In order to show the inaccuracy of these oversimplifications for 131I, results using Monte Carlo radiation transport analysis with simplified anthropomorphic mathematical phantoms were determined. These results were then compared to actual patient measurements and the results of point source analysis. The measurement data were taken from 49 131I radioimmunotherapy patients. The point source calculations were performed using well established methodologies and using the same assumptions as in the NRC regulations for patient release criteria. Monte Carlo results were obtained by implementing two simplified 70 kg anthropomorphic phantoms and performing radiation transport simulation. The activity in the "patient" phantom was assumed to be localized in the abdominal region to correspond to the activity localization seen in the radioimmunotherapy patients who were measured. Dose equivalents per unit cumulated activities were determined for 131I using the various methods. The relationship between measured dose equivalent per unit cumulated activity and whole body dose equivalent per unit cumulated activity was also investigated using Monte Carlo analysis. The point source method as implemented by the NRC yields an estimated dose equivalent per unit cumulated activity of 1.6 x 10(-8) mSv MBq(-1) s(-1) at 1 m (2.2 x 10(-4) rem mCi(-1) h(-1) at 1 m), and the Monte Carlo based method yielded a whole body dose equivalent per unit cumulated activity in the target phantom of 6.8 x 10(-9) mSv MBq(-1) s(-1)(9.0 x 10(-5) rem mCi(-1) h(-1)) for abdominal localization of activity in the source phantom. The measurements of the radioimmunotherapy patients yielded an average result of 1.0 x 10(-8) mSv MBq(-1) s(-1) (13 x 10(-4) rem mCi(-1) h(-1)). When corrected for the difference between measured surface dose equivalent and whole body dose equivalent as determined by Monte Carlo analysis, these measurements represent a whole body dose equivalent per unit cumulated activity of about 6.2 x 10(-9) mSv MBq(-1) s(-1) (8.1 x 10(-5) rem mCi(-1) h(-1)). Based on these results, the current NRC dose-based methodology for the release of patients administered radioactive materials significantly overestimates the dose equivalent to others from 131I therapy patients.