This talk will discuss a new model-based clustering methodology. This approach is predicated on the use of a recursive semi-parametric density estimation procedure, the adaptive mixture method of Priebe, as the starting point to the agglomerative phase of the model-based clustering procedure. The computational efficiency of the model-based clustering methodology is improved in that one does not have to perform the agglomerative phase of the procedure using the complete data set. The methodology along with recent results obtained on artificially generated data will be presented.