The development of techniques to record from populations of neurons has made it possible to ask questions concerning the encoding of task-relevant information in awake, behaving animals. The issue of how groups of neurons within different brain structures register and retrieve representations of behaviorally significant events can now be addressed using multineuron-recording techniques. This review examines recent studies employing simultaneous recording of ten or more individual neurons in the mammalian brain. A major issue discussed is whether ensemble information content reconstructed from single-neuron recordings may be underestimated if compared to ensembles where those same neurons were recorded simultaneously. The mechanics of ensemble information encoding in the hippocampus is illustrated from population statistical analyses of ensemble activity during performance of a delay task. Detailed descriptions of methods of extracting ensemble information, as well as cross-correlational analyses, are discussed in the context of emergent issues regarding interpretation of ensemble data.