Abstract::The development of clinically viable metformin analogs is a challenge largely to be overcome. Despite being an extremely efficient drug for the treatment of type 2 diabetes mellitus, multiple studies were conducted seeking to improve its hypoglycemic activity or to ameliorate aspects such as the low oral absorption and the incidence of gastrointestinal side effects. Furthermore, efforts have been made to attribute new activities, or even to expand the pre-existing ones, that could enhance its effects in the diabetes, such as pancreas-protective, antioxidant, and anti-inflammatory activities. In this manuscript, we describe the analogs of metformin developed in the last three decades, highlighting the lack of computationally based rational approaches to guide their development. We also discuss this is probably a consequence of how unclear the mechanism of action of the parent drug is and highlight the recent advances towards establishment of the main molecular target(s) for metformin. We also explored the binding of metformin, buformin and phenformin to the mitochondrial respiratory chain complex I through molecular docking analyses and reviewed the prospects of applying computational tools to improve the success in the development of such analogs. Therefore, it becomes evident the wide range of molecular targets, as well as the multiple activities displayed by metformin, make this drug a promising prototype for the development of novel entities, particularly for the treatment of type 2 diabetes mellitus.