Clinical trials involving p53 gene therapy for ovarian cancer failed due to the dominant negative inhibition of wild-type p53 and multiple genetic aberrations in ovarian cancer. To overcome this problem, we have designed a more potent chimeric gene fusion, called p53-Bad, that combines p53 with the mitochondrial pro-apoptotic factor Bad. Unlike wild-type p53, which acts as a nuclear transcription factor, this novel p53-Bad construct has multiple unique mechanisms of action including a direct and rapid apoptotic effect at the mitochondria. The mitochondrial localization, transcription activity, and apoptotic activity of the constructs were tested. The results suggest that p53 can be effectively targeted to the mitochondria by controlling the phosphorylation of pro-apoptotic Bad, which can only localize to the mitochondria when Ser-112 and Ser-136 of Bad are unphosphorylated. By introducing S112A and S136A mutations, p53-Bad fusion cannot be phosphorylated at these two sites and always localizes to the mitochondria. p53-Bad constructs also have superior activity over p53 and Bad alone. The apoptotic activity is consistent in many ovarian cancer cell lines regardless of the endogenous p53 status. Both p53 and the BH3 domain of Bad contribute to the superior activity of p53-Bad. Our data suggests that p53-Bad fusions are capable of inducing apoptosis and should be further pursued for gene therapy for ovarian cancer.