Recent studies have demonstrated that differentiated somatic cells from various mammalian species can be reprogrammed into induced pluripotent stem (iPS) cells by the ectopic expression of four transcription factors that are highly expressed in embryonic stem (ES) cells. The generation of patient-specific iPS cells directly from somatic cells without using oocytes or embryos holds great promise for curing numerous diseases that are currently unresponsive to traditional clinical approaches. However, some recent studies have argued that various iPS cell lines may still retain certain epigenetic memories that are inherited from the somatic cells. Such observations have raised concerns regarding the safety and efficacy of using iPS cell derivatives for clinical applications. Recently, our study demonstrated full pluripotency of mouse iPS cells by tetraploid complementation, indicating that it is possible to obtain fully reprogrammed iPS cells directly from differentiated somatic cells. Therefore, we propose in this review that further comprehensive studies of both mouse and human iPS cells are required so that additional information will be available for evaluating the quality of human iPS cells.