predicting sea surface temperature (SST) and atmospheric pressure anomalies in the eastern Pacific, culminating in the early, widely publicized forecast of the 1997-98 ENSO event. Now ENSOmania has swept the globe. Agricultural and resource economists have a special interest in ICF because these sectors are especially affected by climatic fluctuations. In commenting on the three papers in this session, I make five interrelated points about the economics of climate forecasting. First, El Nifio is heavily overhyped because forecast skill with respect to ENSO phase is not the same as the ability to forecast ENSO's impacts on variables of direct interest to decision makers. The papers by Mjelde, Hill, and Griffiths (MHG) and by Podbury, Sheales, Hussain, and Fisher (PSHF) explore different dimensions of this point. Forecasts have value only if they alter decision makers' subjective probability distributions over stochastic variables that influence choice. Few people care directly about SSTs or atmospheric pressure over the eastern equatorial Pacific (i.e., ENSO events). Even among Peruvian fishermen, interest in ENSO is indirect in its imperfectly understood effects on the spatial and temporal distribution of fish populations. For most people, ENSO forecasts are mediated by predicted climate teleconnections, the relationship between local climate and distant ocean temperature and pressure anomalies. Such links are statistically strong in relatively few places and only at coarse spatial and temporal resolution. Moreover, few choices are based directly on climate. Rather, climate has indirect effects on decisions through its relation to crop yields, livestock productivity, food prices, stream flow, and so on. Because ENSO forecast information is only of indirect interest to decision makers, its value is conditioned by the robustness of the mediating inferences, with forecast errors compounding in a complex fashion to the point where they might not change subjective probability distributions.