Consider the following experiment: At stress levels x1< … < xk, n1, …, nkitems are put on test. The failure times of those items failing before timest1*,tk* are recorded, and the remaining failure times are censored. LetP(t,x) denote the probability that an item stressed at levelxfails on or before timet. We consider models of the formP(t,x) = Φ[g(t,x)], where g is a real-valued function specified up to an unknown parameterc, and Φ is an unspecified continuous distribution function. Under these assumptions a Kolmogorov-Smirnov-type test is discussed as a method for estimating c as well asP(to, xo) for specifiedtoand xo. The method is exact and does not require failures at unaccelerated stress levels. Ifn1= … =nk, the method extends to experiments that at each stress level are terminated after a specified number of failures is observed.