Saraswat, P. and Sen, M.K., 2012. Pre-stack inversion of angle gathers using a hybrid evolutionary algorithm. Journal of Seismic Exploration, 21: 177-200. Inversion of pre-and post-stack seismic data for acoustic and shear impedances is a highly non-linear and ill-posed problem. A deterministic inversion of band-limited seismic data produces smooth models that are devoid of high frequency variations observed in well logs. The objective of this paper is two-fold, i.e., to develop an efficient scheme to explore and exploit the model space, and to efficiently sample broadband models statistically. We demonstrate that the use of starting models from fractal based a priori pdfs helps us to derive elastic models of very high resolution. We also introduce a new hybrid inversion algorithm that takes advantage of both deterministic and stochastic methodologies. A deterministic inversion based on conjugate gradient (CG) method produces smooth models while a stand-alone stochastic method based on differential evolution (DE) produces high-resolution models of nearly the same accuracy. A hybrid algorithm that uses CG solution as a starting model converges much faster than a standalone DE to very good solutions. We demonstrate our results with application to a field seismic dataset. The hybrid algorithm can also be used to sample the most significant parts of the model space rapidly resulting in estimates of uncertainty.