The purpose of this report is to compare two long-run forecasts of participation in fishing, hunting, and nonconsumptive wildlife recreation based on the 1980 and 1985 national surveys. The study addresses the question of stability to the empirical relationships and long-run predictability of the logit model. The research procedure follows federal guidelines recommending that forecasts of recreation demand be based on multiple regression analysis which provides coefficients estimating how much each of the explanatory variables causes participation to vary. Participation in both years is shown to be a function of population, a travel cost proxy for price, the price of substitutes, income, age, residence, and other socioeconomic characteristics of individuals, quality of the experience, and availability or resources. Both forecasts are based on predicted changes in these exogenous variables available from the U.S. Census and other published sources. Indications are that nonconsumptive wildlife recreation will be the fastest growing activity.