Much of the U.S. progress in short term climate forecasting has been as a result of a concerted and focussed effort. The TOGA Program on Prediction (T-POP) was instituted in the U.S. in 1991 as a distributed research program and was designed to develop prediction systems, to develop methods of data assimilation for analysis and initialization, and to do the modeling and theoretical research needed to delineate and advance the theory of predictability of the climate on seasonal-to-interannual time scales [see Cane and Sarachik, 1991]. T-POP has provided a forum for the development of short term climate prediction systems and has instituted an intercomparison of forecasts by different groups for the common period 1985-1990. Under NOAA funding, it supports all the seasonal-to-interannual forecast development activities in the United States.