The observations on patients treated with neutrons from the 60 inch Berkeley cyclotron between 1938 and 1943 are reviewed and reinterpreted. This investigation is concerned only with the effects of neutrons on skin and subcutaneous tissues. Utilizing data that relate the relative biological effectiveness (RBE) for neutrons to the size of the dose-fraction, the neutron doses were converted to equivalent megavoltage roentgen doses. When applicable, factors were introduced to account for the contribution from exit beams. The timedose-fractionation pattern varied from patient to patient and usually was inconsistent with schemes used in current radiation therapy. Therefore, the doses were calculated in terms of a nominal single dose (NSD) and a nominal 6 week dose (SWD). The NSD and SWD allow some comparison with conventional radiation therapy. Suitable for inclusion were 144 patients with 275 separate fields treated with the 60 inch cyclotron. Among this group were 25 survivors, with 51 treatment fields, who were observed 2 years or more after neutron therapy was completed. Early reactions were judged according to findings during the first 2 months after treatment. The late reactions were graded according to observations 2 or more years after therapy. The data indicate that: (a) the roentgenray equivalent of the neutron doses applied was relatively large in comparison with conventional megavoltage roentgen-ray therapy; (b) an early reaction was visible in all treatment fields and in 60 per cent this amounted to epidermolysis; (c) the late reactions were severe in most patients who survived 2 years or longer, but in 80 per cent of the fields epidermolysis had developed at the time of treatment; (d) fields in which late ulceration developed had received SWDs of 6,100 to 11,000 (mean 7,800 ± 500 rads); (e) tissues with a high content of fat appeared to have a greater reaction; (f) the "calculated" tolerance SWD for this neutron trial is about 6,100 rads; and (g) after the tolerance dose is exceeded, the severity of the late reaction appears to increase more rapidly as a function of dose than does the severity of the early reaction, but this may be related to the grading schemes used. With the proper allowance for exit dose, fractionation scheme, and change of RBE with fraction size, both early and late skin reactions can be accounted for on the basis of dose received. Furthermore, the late reactions seem generally consistent with the early reactions. Lack of knowledge regarding (a) the influence of the fractionation scheme on tissue response, (b) possible skin-sparing with fast neutrons, and (c), as suggested by the Hammersmith group, the change of RBE with fraction size may have caused the earlier workers to underestimate doses applied and to misinterpret the significance of the early skin reactions. We believe that the Berkeley neutron data from 1938 to 1943 should not contraindicate a properly planned and controlled clinical investigation of neutron therapy.