Mixture models are frequently employed in astronomical studies to model observed data and interpret results. Gaussian mixture model (GMM) is probably the most widely used one due to its simplicity. To illustrate, GMM had been applied to the pulsar data set in a previous study and discovered six clusters. On the other hand, there are more sophisticated mixture models e.g. Dirichlet process Gaussian mixture model (DPGMM). It is a Bayesian non-parametric model such that it includes prior distributions for model parameters and automatically explores the optimum number of clusters in a data set, in contrast to GMM. In this study, we repeated the application of GMM, and also tested DPGMM as a first time on a larger pulsar data set. It is revealed that there arc six clusters in the data set as presented in the former study, according to both GMM and DPGMM. However, the estimated parameters of both models differ from each other. We, then, compared the clustering performance of models with respect to silhouette coefficients. Accordingly, it is observed that DPGMM exhibits better clustering performance. As a further analysis, we compared the classification performance of models. Apparently, DPGMM performs, once again, better than GMM in discriminating selected pulsar families.