Risk-based allocation strategies, also known as Smart Beta allocations, define the weights of assets in portfolios as functions of the individual and common asset risk. In this paper we focus on the Minimum Variance (MV), Maximum Diversification (MD), Equal Risk Contribution (ERC) and Equal-Weight (EW) risk-based allocation strategies. The popularity of risk-based strategy is commonly justified by their good record of out-performing the cap-weighted (CW) allocation strategy. Because of the low-volatility profile of risk-based allocations this is especially true when crises occur. From March 15, 2002 to May 1, 2012 we investigate how using a socially responsible investment universe impacts performance of risk-based allocation strategies. We use different measures of performance, included risk-adjusted one (multi-factor models), and we propose to disentangle the effect of using a SRI universe from the effect of using risk-based allocations. SRI universe only contains firms that have good environmental, social and governance performance. This kind of filtering is increasingly popular among institutional investors. On the estimation period, using European stocks, we find that the use of the SRI universe has a positive contribution to risk-adjusted performance of risk-based allocations. However this contribution is not uniform among all the risk-based allocation strategies and, can represent only a small part of the total alpha that is observed.