An instrument, having great commercial potential, has been developed for monitoring chemical processes using a passive acoustic technique,The instrument has been developed with the particular target of monitoring crystallisation reactions. Pattern recognition and neural network techniques can be used to improve the performance of the monitoring technique, using the acoustic data to produce an estimate of process parameters which cannot otherwise be easily measured. The technique has been shown to be capable of estimating the size of suspended solids in an agitated vessel, and estimating the yield from a crystalliser. However, for reliable results, care must be taken in acquiring the training data for the neural network based pattern recognition algorithm.