Cooperative diversity can be applied to energy-constrained wireless sensor networks to significantly reduce node energy consumption. However, cooperation partners must be carefully selected and coordinated to practically exploit this energy saving potential. In this paper we investigate partner choice for energy efficient cooperation in a wireless sensor network. We formulate novel and computationally efficient partner choice heuristics for sensor nodes based on either global or local knowledge of average path loss values in the network. We present extensive simulation results of cooperation in a wireless sensor network to show that the proposed heuristics achieve near-optimally energy efficient partner selection. Our results also demonstrate that large network-wide energy savings are achieved as a result of cooperative communication. Therefore, our simple partner choice heuristics form the basis of an effective distributed cooperation protocol for improving the energy efficiency of a wireless sensor network. Very importantly from the point of view of practical implementation, we show that our partner choice heuristic based on local information is the most effective cooperation strategy for resource-constrained wireless sensor networks, as it yields superior energy conservation results while enabling fully distributed and scalable cooperation.