The paradigm states that cytoplasmic actin operates as filaments and nuclear actin is mainly monomeric, acting as a scaffold in transcription complexes. However, why should a powerful function of actin, namely polymerization, not be used in the nucleus? Recent progress in the field forces us to rethink this issue, as many actin filament assembly proteins have been linked to nuclear functions and new experimental approaches have provided the first direct visualizations of polymerized nuclear actin.