This paper presents an adaptive algorithm that improves the energy efficiency of VoIP applications over IEEE 802.11 networks. The algorithm seeks to achieve the largest energy savings subject to reaching a minimum speech quality under the prevailing network conditions. The control mechanism used is the dynamic selection of the packet size during the communication. Three variants of the control algorithm are presented and compared.Our proposal has been implemented in a VoIP application and experimentally validated in a testbed, and the results demonstrate that our packetization period control algorithm can provide energy savings in uncongested IEEE 802.11 networks (up to 30%). Furthermore, under poor network conditions, the algorithm can prolong the duration of the call before it is dropped at the expense of higher energy consumption during such conditions. Each variant of the algorithm provides a unique balance between keeping quality and saving energy.