El Niño events—anomalous warmings of the tropical Pacific with associated climatic and economic impacts around the globe—have occurred at several‐year intervals since before written records began with the logs of Francisco Pizarro in 1525. In this review, the history of El Niño research is traced from its beginnings through the key innovations of Bjerknes and Wyrtki to the unusual 1982–1983 event. Recent research is then reviewed, with detailed discussions of two important processes: instability growth and vacillation between climate states. Throughout the paper there are adjunct discussions of extraregional teleconnections, ecological impacts, and research on El Niño in the ancient record. The final section discusses the present paradigm for vacillations between El Niño and non‐El Niño states and speculates on the possibly chaotic nature of El Niño. El Niño and its atmospheric counterpart, the Southern Oscillation, appear to occur as an internal cycle of positive and negative feedbacks within the coupled ocean‐atmosphere climate system of the tropical Pacific, although hypotheses based on external forcing also exist. All events are preceded by westerly wind anomalies on the equator near the date line. Baroclinic equatorial Kelvin waves are generated, propagating eastward toward South America where they depress the thermocline and raise sea level, while the deep, upper ocean reservoir of warm water in the western Pacific is depleted. Sea surface temperature (SST) anomalies in the cool eastern Pacific occur primarily because the normal source of cold water is depressed below the reach of mixing and upwelling processes. In the central equatorial Pacific, eastward advection by anomalous zonal flows is the principal mechanism. Nonlinear heat transfer to the lower atmosphere creates a positive ocean‐atmosphere feedback resulting in the unstable growth of anomalies along the equator. Much of the present research aims at determining how the ocean‐atmosphere system vacillates between the El Niño and non‐El Niño states. Coupled models suggest that a longer time scale, negative‐feedback process produces the transitions: at the apex of an El Niño development an anomalous atmospheric convection above the areas of maximum SST produces areas of reduced upper layer thickness in the off‐equatorial ocean, which slowly propagate westward to the western boundary as Rossby waves and back to the central equatorial Pacific as upwelling Kelvin waves, reestablishing the normal cooling process. A similar negative feedback of opposite sign completes the second half of an oscillation, returning again to the El Niño state. However, the notion that El Niño‐Southern Oscillation variability results only from an internal feedback process is still highly contentious, and a number of external fo