In this paper, we study an adaptive random search method based on continuous action-set learning automaton for solving stochastic optimization problems in which only the noisecorrupted value of function at any chosen point in the parameter space is available. We first introduce a new continuous action-set learning automaton (CALA) and study its convergence properties. Then we give an algorithm for optimizing an unknown function. r 2005 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.