An iterative reconstruction algorithm for sodium magnetic resonance imaging (23Na MRI) with multi-channel receiver coils is implemented and compared to a conventional gridding reconstruction. Based on compressed sensing (CS) it utilizes a total variation (TV(2)), combined with anatomical weighting factors (AnaWeTV(2)) to preserve known tissue boundaries. Simulated and measured 23Na multi-channel data sets of the female breast were reconstructed. The TV(2) and in particular the AnaWeTV(2) lead to an improved image quality, due to effective noise reduction and the highlighting of structure. The presented CS reconstruction is beneficial especially for high undersampling factors.