Since the energy budget of mobile nodes is limited, the performance of a networking protocol for such users should be evaluated in terms of its energy efficiency, in addition to the more traditional metrics such as throughput. In this paper, two topology-unaware MAC protocols - in which the scheduling time slots are allocated irrespectively of the underline topology - are considered and their energy consumption is derived. It turns out that the per frame power consumption is lower for the less throughput-efficient protocol, suggesting that energy savings are achieved at the expense of throughput. A finer energy consumption study is carried out in the sequel, focusing on the amount of energy consumed to successfully transmit a certain number of packets, or equivalently, on the per successful transmission power consumption. It is shown that the more throughput-efficient protocol consumes less energy per successful transmission under certain conditions (which are derived), due to the low...