Surgical gloves were manufactured using the RVNRL process. A fractional factorial design at two levels showed that five parameters of the coagulant dipping process which were studied independent. Coagulant concentration and dwell time in the radiovulcanized latex presented major main effects while the temperature of the former before dipping into the radiovulcanized latex and the flow time of the radiovulcanized latex on the former surface presented opposite main effects. The withdrawal rate of the former from the radiovulcanized latex did not change glove thickness. The mathematical correlation between the estimates of thickness and the significant main effects of coded variables was y = 0.212+0.025x1+0.019x2. This optimized equation allowed reproduction of a surgical glove thickness in the range of 0.157 to 0.291 mm, which is considered acceptable by international standard specification.