One of the most popular feature extraction algorithms for Brain-Computer Interfaces (BCI) is the Common Spatial Patterns (CSP) algorithm. Despite its known efficiency and widespread use, CSP is also known to be very sensitive to noise and prone to overfitting. To address this issue, some groups have recently proposed to regularize CSP. In this paper, we present a simple and unifying theoretical framework to perform such a CSP regularization. We then present a mini-review of existing Regularized CSP (RCSP) algorithms, and describe how to cast them in this framework. We also propose 4 new RCSP algorithms. Finally, we compare the performances of 11 different RCSP algorithms (including these 4 new ones and the original CSP), on EEG data from 17 subjects, from BCI competition data sets. Results showed that the best RCSP methods can outperform CSP by nearly 10% in median classification accuracy and lead to more neurophysiologically relevant spatial filters. They also enable us to perform efficient subject-to-subject transfer. Overall, the best RCSP algorithms on these data were the CSP with Tikhonov Regularization and Weighted Tikhonov Regularization, both newly proposed in this paper.