By and large, the agrarian based countries of the tropics are at the mercy of climate variability and skillful predictions of this variability mean a great deal to these countries. There is sufficient skill demonstrated in the existing forecasts that these forecasts are already being used for agricultural planning in Peru [ Lagos and Buizer, 1992]. The rainy season, and therefore the growing season, in coastal northwest Peru is January to April and is primarily determined by the seasonal march of the eastern Pacific Intertropical Convergence Zone (ITCZ). Precipitation anomalies in this region are known to correlate extremely well with the SST anomalies characteristic of ENSO. By law, Peru must produce an agricultural plan each year, which determines the allocation of crops, seeds, water, power, etc. This plan is heavily influenced by existing ENSO predictions: a prediction of warm and wet implies a shift to rice cultivation while a prediction of cold and dry shifts the allocations to cotton. The prediction need only be available a few months before the start of the growing season and, since the ENSOs in the last decade have been of the type that grow first in the equatorial Pacific and then in coastal South America, these predictions have been successful and have succeeded in increasing total agricultural output.