Mobile agents (MAs) are referred to as autonomous application programs with the inherent ability to move from node to node towards a goal completion. In the context of wireless sensor networks (WSNs), MAs may be used by network administrators in the process of combining data and knowledge from different sources aiming at maximizing the useful information content. MAs have been initially developed to replace the client/server model which exhibits many disadvantages, particularly in WSN environments (e.g.heavy bandwidth usage and excessive energy expenditure). The most promising advantages of MAs in WSN environments include decreased usage of the wireless spectrum (large volumes of raw sensory data are filtered at the source) and energy consumption, increased reliability due to their inherent support for disconnected operations, their ability of cloning themselves to enable parallel execution of similar tasks, etc. The main objective of this chapter is to review and evaluate the most representative MA-based middleware proposals for autonomic data fusion tasks in WSNs and evaluate their relevant strengths and shortcomings. Although the chapter’s focus is on autonomic data fusion tasks, other applications fields that may benefit from the MAs distributed computing paradigm are identified. Open research issues in this field are also discussed.