Data obtained in reactor dosimetry, particularly from reactors exclusively serving for irradiation, can today adequately treated making use of the MCNP or other code of the Monte Carlo method, although the need to prepare the benchmark problem to validate the data for Monte Carlo treatment will limit the range of experiments. The present paper discusses the usefulness and effectiveness of the Monte Carlo method, citing typical problem treated by this means in the materials testing reactor JMTR, the experimental fast reactor "JOYO" and PWRs/BWRs in general.