ABSTRACTThe potential to cause a primary irritant reaction is usually evaluated by clinically recognizable dermatitis at the site of patch testing. Such evaluation is difficult in the case of substances with low irritancy potential which sometimes fail to cause grossly visible signs of dermatitis. In those cases a higher concentration of the test substance may be used in order to produce a visible reaction, but such concentrations may lead to unnecessary irritation and are of little value in determining the safety of the substance in concentrations of practical use.The microscopic appearance of normal human skin observed on a skin replica made by Susuki's Universal Microscopic Printing (SUMP method) has been reported previously by the author (9). In the present study, patch test reactions which fail to produce grossly visible inflammation were observed for microscopic changes by the SUMP method. The patches were applied to the upper arms of normal humans for 24 hours and then the test sites were observed for gross inflammation and microscopically by the SUMP method. Numerical scores were assigned to grade each microscopic change produced in the skin surface and total scores were used to grade the irritancy reaction produced. These scores were compared to results using Zontes TA‐430, which has been used for over 10 years as a softening agent of fibers and was used in this study as a control chemical substance with no irritancy. The results revealed microscopic changes provoked by irritants too weak to produce gross inflammation. A criteria for grading cutaneous primary irritant reactions by microscopic findings was determine