In wireless sensor networks (WSNs), energy consumption and data quality are two important issues due to limited energy resources and the need for accurate data. In this scenario, information fusion techniques (also reffered to as data fusion or data aggregation) have been used to distributedly process sensor data to save energy ? by reducing the overall communication load ? and increase accuracy ? by removing noisy measurements of correlated data. The thesis herein summarized presents a study about how information fusion techniques can be used in WSNs to improve data accuracy and reduce the energy consumption. The main contributions include: (1) an information-fusion framework for WSNs; (2) a novel algorithm that applies information fusion to detect when a routing tree needs to be rebuilt; (3) a novel routing strategy, based on role assingment, which maximizes the gains of an information-fusion application; and (4) a critical survey about information fusion in WSNs.