A method for exploiting human's internal contamination data for radioactive release estimation in nuclear power plant accidents is proposed. Nevertheless, such data is often very rough and uncertain; it is accessible even in toughest situations when most of the active and passive monitors are damaged by the accident. These data can be used in combination with other collectable data for estimating the event scale in severe nuclear power plan accidents. The rationale behind the method is that nuclear power plant accidents are often associated with internal contamination of radiation workers involved in the early stages of emergency response activities mainly due to the release of 131I in atmosphere. The proposed inverse analytical approach uses the 131I intake of contaminated workers, their working conditions, chronology of events, and applied personal safety measures during the first hours or days of the emergency response activities to estimate the magnitude of 131I concentration in the air