A new methodology is introduced and empirically tested for estimating the economic benefits accruing to society from an improved recreational facility. The specific facility under consideration is Upper Klamath Lake, Oregon, which presently has low water quality. The methodology draws upon previous work done in the evaluation of recreational demand; however, it focuses upon the individual recreationist and separates the traditional price variable into on-site costs and travel costs. The model is used to estimate the number of days per visit the recreationist will stay at the site as the water quality improves. Data collected at three other lakes with varied characteristics are used to derive a relationship between the number of visits to a site and the characteristics of the site. The impact of expanded recreational use of Klamath Lake upon the local economy is also estimated through the use of an input-output model of the Klamath County economy. (Author Modified Abstract)