The natural reservoirs for avian influenza (AI) viruses are wild aquatic bird species of the orders Anseriformes and Charadriiformes. Infected birds readily excrete virus in their feces contaminating the surrounding aquatic environment with transmission occurring via the fecal-oral route. Previous studies have established that AI viruses can remain infectious in water for several weeks to months, suggesting that surface water may serve as both a transmission medium and potentially a long term reservoir. The majority of the studies focused on AI virus persistence in water have utilized sterile distilled water to simulate the surface water of aquatic ecosystems. The objective of this research was to identify and characterize the factors that affect the duration of AI virus persistence in natural surface water.The duration of infectivity was determined for two common AI virus subtypes in 15 filtered surface water samples collected from waterfowl habitats in Georgia. Consistent with previous studies, viruses were less stable at warmer temperatures and in acidic water (pH<5). Variations in persistence times were observed between water samples with comparable pH and salinities indicating that other abiotic factors affect the stability of the virus in natural surface water. To identify the other important abiotic factors, viral persistence trials were performed in filtered surface water from 38 waterfowl habitats distributed across the United States and in-depth chemical analyses were performed. In addition to previously identified abiotic factors, the ammonia, chloride, and sulfate concentrations of surface water were determined to be significant predictors of virus persistence. To evaluate the effect of the biological component of surface water and examine virus strain-related variation in persistence, the loss of infectivity was estimated for nine wild duck isolated AI viruses in three types of water: distilled, filtered surface water, and intact surface water. All viruses persisted longest in distilled water followed by filtered surface water with markedly reduced durations of persistence observed in the intact surface water, suggesting that surface water may not readily facilitate the long term maintenance of AI viruses. These results contribute to our understanding of the role of surface water in the ecology of AI virus.