This dissertation has two main objectives: compare the performance of equity European Socially Responsible Funds (SRFs) with the market benchmark; analyse how the performance of SRFs is affected by the use of different filters (Environmental/Social and Negative/Positive). To evaluate fund performance, we use the total conditional model of Christopherson, Ferson and Glassman (1998) applied to the 1-factor model and to the 4-factor Carhart (1997) model. The sample is composed by 111 global and international equity funds from sixteen European countries for the period from January 2000 to October 2014. The performance of the funds is estimated taking into account both a conventional benchmark (MSCI AC ALL WORLD E) and a socially responsible benchmark (FTSE4GOOD GLOBAL). The results show us that, on average, SRFs have a similar performance comparing with the market. However, when we use the socially responsible benchmark, the performance of these funds is slightly higher. The explanatory power is, on average a little bit higher if we consider all the sample, when we use the conventional benchmark (both for the 1-factor model and 4-factor model). On a country basis analysis, we can conclude that only half of them present a higher coefficient of determination. With regard to size, value and momentum factors, SRFs seem to be largely exposed to small companies. And in a less extent, also more exposed to growth stocks than to value stocks. In relation to the momentum factor, the results are not significant. The results also show little evidence of time-varying alphas but there is a clear evidence of time-varying betas. This is even more evident when we use the conditional 4-factor model with the socially responsible benchmark. When analysing per filters, we conclude that the performance of SRFs is not affected by the type of social filters used.