We document the nature and frequency of seal predation at Atlantic salmon (Salmo salar) farms in Maine and determine whether the severity of predation is related to the proximity of farms from one another and nearby harbor seal (Phoca vitulina concolor) haul-outs. We surveyed farm managers annually from 2001–2003 to document management techniques, husbandry practices, and predator deterrence methods employed for comparison with the extent of seal predation. Biweekly aerial surveys were conducted between January and March of each year to document harbor seal presence. An empirical estimate from a negative binomial model showed seal predation at farms declined significantly with distance to the nearest haul-out, suggesting that seal predation may be deterred by maximizing the distance between farms and seal haul-outs. Farms located further than 4 km from harbor seal haul-outs experienced minimal losses. At farms located within 4 km of harbor seal haul-outs, seal predation decreased with increasing distance from neighboring farms, indicating that areas where farms are concentrated may be more vulnerable. The regular replacement of primary and secondary cage netting was negatively correlated with seal predation. Finally, this study documents the apparent ineffectiveness of acoustic harassment devices at deterring seal predation.