An experimental approach was used to investigate wintering waterfowl responses to introduced trail use at foraging sites with and without recreational trails in California, USA. Waterfowl were exposed to trail use in the form of two researchers walking levees adjacent to ponded habitat, and the number of waterfowl by species were compared before and after experimental walks in 40-m bands starting at the levee and extending 200 m into the ponds. The researchers recorded distances to the nearest individuals, responses of focal animals, and numbers of recreational trail users. The most numerous species were Ruddy Duck (Oxyura jamaicensis), Northern Shoveler (Anas clypeata), and scaup spp. (Aythya affinis and A. marila). Recreational trail use rates at trail sites averaged 1 to 82 people/hr. The greatest difference in numbers of birds before vs. after experimental walks occurred in the two 40-m bands closest to the levee at non-trail sites (0–40 m [t = 4.558, P = 0.0001], 40–80 m [t = 3.775, P = 0.001]) and trail sites (0–40 m [t = 3.049, P = 0.005], 40–80 m [t = 1.808, P = 0.082]). The relationship between the ratio of beforeto after-walk waterfowl numbers vs. date since the start of the winter season (r² = 0.315, P = 0.102) and the total number of birds vs. the number of recreational trail users (r² = 0.041, P = 0.847) did not indicate increasing tolerance to trail use for waterfowl overall. However, species varied in their tolerances. Ruddy Duck numbers declined with increasing numbers of recreational trail users (rho = -0.481, P = 0.017), while Northern Shoveler numbers increased (rho = 0.456, P = 0.025). Distances (using the 95th percentile) that individual birds were recorded from researchers during experimental walks varied from approximately 170–200 m at both non-trail and trail sites.