This article outlines a methodology for estimating the number of individual visitors to a set of recreational sites, as well as counts of specific visitor groups. The model is designed for sets of sites characterized by: (a) unrestricted recreation from a wide and partially unknown geographic market; (b) individuals who may visit more than one site; (c) accurate visitation (gate) counts from each site. The model provides consistent estimates of the total number of individual visitors, based on information embedded in site-level count and survey data. Monte Carlo analysis and an empirical application illustrate the properties of visitor count estimates. Copyright 2003, Oxford University Press.