We fitted Northern Pintail Anas acuta in Japan with satellite transmitters and monitored their spring migration movements relative to locations where the highly pathogenic H5N1 avian influenza virus was detected in Whooper Swans Cygnus cygnus in 2008. Pintails were assumed not to be infected with the H5N1 virus at the time they were marked because capture occurred between 2 and 5 months before reported outbreaks of the virus in Japan. We assessed spatial and temporal overlap between marked birds and occurrence of the virus and tracked Pintails after they departed outbreak locations. Eight of 66 (12.1%) Northern Pintails marked with satellite transmitters used wetlands in Japan where the H5N1 virus was detected in Whooper Swans. Apparent survival did not differ between Pintails that used H5N1 sites and those that did not. However, the proportion of Pintails that migrated from Japan was significantly lower among birds that used H5N1 sites compared with those that did not (0.50 vs. 0.79). Northern Pintails were present at the H5N1 sites from 1 to 88 days, with five birds present at the sites from 0 to 7 days prior to detection of the virus in Swans. The six Pintails observed to depart H5N1 sites did so within 2–77 days of the reported outbreaks and moved between 6 and 1200 km within 4 days of departure. Four Pintails migrated to eastern Russia. After their departure from outbreak sites, Northern Pintails made long‐distance migrations within the period when newly infected ducks would shed the H5N1 virus. This supports a hypothesized mechanism by which a highly pathogenic avian influenza virus could be spread by migratory birds.