The working times of five poly(vinyl siloxane) impression materials were estimated by evaluating the dimensional accuracy of stone dies of impressions of a standard model made at successive time intervals. The stainless steel standard model was represented by two abutments having known distances between landmarks in three dimensions. Three dimensions in the x-, y-, and z-axes of the stone dies were measured with a traveling microscope. A time interval was rejected as being within the working time if the percentage change of the resultant dies, in any dimension, was statistically different from those measured from stone dies from previous time intervals. The absolute dimensions of those dies from the rejected time interval also must have exceeded all those from previous time intervals. Results showed that the working times estimated that this method generally were about 30 seconds longer than those recommended by the manufacturers.