Monitoring population trends in harbor seals (Phoca vitulina) generally involves two steps: (i) a census obtained from aerial surveys of haul-out sites, and (ii) an upward correction based on the proportion of seals hauled out as estimated from a sample of telemetry-tagged seals. Here we present a mathematical method for obtaining site-specific correction factors without telemetry. The method also determines site-specific environmental factors associated with haulout and provides algebraic equations that predict diurnal haul-out numbers and correction factors as functions of these variables. We applied the method at a haul-out site on Protection Island, Washington, USA. The haul-out model and correction factor model were functions of tide height, current velocity, and time of day, and the haul-out model explained 46% of the observed variability in diurnal haul-out dynamics. Although the particular models are site-specific, the general model and methods are portable. A suite of such models for haul-out sites of a regional stock would allow managers to monitor long-term population trends without telemetry.