Sensor networks are characterized by limited battery supplies. Due to this feature, communication protocols specifically designed for these networks should be aimed at minimizing energy consumption. To this purpose, the sensor's capability of transmitting with different power levels can be exploited. With this in mind, in this paper an integrated MAC/routing protocol, called MACRO, which exploits the capability of sensor devices to tune their transmission power is introduced. The proposed protocol requires that each node only knows its own coordinates and the coordinates of the destination, but does not require any exchange of location information. In order to select the next relay node, a competition is triggered at each hop, so that the most energy efficient relay node is chosen. This is achieved through maximization of a newly introduced parameter, called weighted progress factor, which represents the progress towards the destination per unit of transmitted power. To this aim, an analytical framework which guarantees that MACRO performs the best choice is derived. MACRO performance is evaluated through ns-2 simulation and compared to other relevant routing schemes. Performance results show that the proposed protocol outperforms other solutions in terms of energy efficiency and boosts data aggregation.