Geoacoustic inversion is the problem of estimating the acoustic parameters of the ocean bottom from acoustic field measurements in water. These parameters may be estimated by the method of matched field processing (MFP), which involves a search in a multi-dimensional parameter space for locating the global maximum of an ambiguity function. Any realistic acoustic model of the ocean bottom involves about 20 geoacoustic parameters and an exhaustive search in such a high-dimensional space is a near-impossible task. In this talk, the application of two stochastic search techniques viz. Genetic Algorithm and Simulated Annealing for solving this global optimisation problem with be discussed. Simulation results will be presented to illustrate the performance of these techniques and some future directions of work for application of these or similar techniques to real data will be discussed.