Harry Wexler was a close student of developing techniques in the field of long range forecasting, and a contributor to them, realizing that improvement of long range forecasts represents one of the most important goals of meteorology. His scientifically-oriented administrative ability played a large part in enlarging the role of this subject in the United States. Indeed, one of the principal reasons he pioneered in inaugurating the World Weather Watch program was to lay a firmer foundation for long-range forecasting. I was one of many who gained benefit and stimulation from a long friendship with him, and my lecture will recall some of the instances. First, I want to describe the nature of the long range prediction problem as seen through the eyes of a pragmatist, and to present a balanced picture of what is now possible and what may become possible in the next decade or so. One might hope that this 10-year forecast of “weather predictability” will turn out better than a 10-year forecast of the weather itself! Next, I will speak on the history of long range forecasting over the past century, as traced through the work of inventors of synoptic, statistical and physical approaches to the problem. In spite of decades of frustration imposed by lack of adequate data, ignorance of large-scale physical processes, and the absence of intensive and large-scale effort, meteorologists working on longe range problems have made encouraging progress.