Marine and off-shore components including propellers, pumps, valves, pipelines and other submerged surfaces are subjected to severe degradation by erosion. Impingement of solid particles mixed in a liquid, referred as slurry, leads to significant material loss and shortens the life span of components. Tailoring the surface properties of materials is an economical way for addressing their degradation. Surface modification through high strain-rate deformation is widely used to enhance functional properties of materials. However, surface modification, particularly at low temperature, is extremely challenging for high strength materials such as stainless steel and has not been investigated comprehensively so far. In the present work, high strain-rate deformation of austenitic steel, SS316L, was performed by innovative submerged friction stir processing technique. For comparative studies, friction stir processing was also performed under ambient cooling conditions. Electron back scatter diffraction studies showed significant grain refinement for the sample processed under submerged conditions. The erosion behavior of as-received and processed steel was investigated using slurry erosion tests. Erosion tests were performed at constant impact velocity of 20 m/s and particle size, while varying the impingement angles. The sample processed under submerged conditions showed nearly two times higher erosion resistance compared to as-received steel. The enhancement in erosion resistance is explained using structural rejuvenation achieved at high strain-rate deformation. All the samples showed similar erosion mechanisms with micro-cutting and ploughing being evident at acute angles and platelet mechanism at normal impingement angle. Erosion phenomena showed a strong correlation with material's hardness at oblique impingement angle while, erosion behavior at normal impingement is explained by the flow work given as hardness to elastic modulus ratio. The study provides fundamental insights into material design for advanced structural applications. © 2017 Elsevier B.V.