Estimates of methane (CH4) emissions from slurry tanks or lagoons in the past have been made primarily by using chamber techniques, which are point specific and interfere with conditions at the slurry-atmosphere interface. This study is based on the use of sulphur hexafluoride (SF6) as a tracer gas to estimate CH4 transport from the slurry surface. The tracer was released from the surface of the swine slurry, and air samples were taken at the downwind rim of the tank over a 165 day period from June 12 to November 20, 1995, at the McGill University-Macdonald Campus Farm, Ste-Anne-de-Bellevue, Quebec, and analyzed for CH4 and SF6 concentrations (C). Knowing the SF6 source strength (Q), CH4 source strength was then determined from the measured downwind concentrations using the C/Q ratio. Using this method, annual CH4 emission from the swine slurry tank was estimated at 56.5 kg CH4 m-2 tank surface yr-1 (¿20%). Assuming that diffusion processes at the nearby dairy tank were similar to those at the swine tank, the same C/Q ratio was used to determine CH4 emissions from the dairy slurry based on downwind CH4 concentrations measured over the same period as at the swine tank. Annual methane emission from the dairy tank was estimated at 74 kg CH4 m-2 tank surface yr-1 (¿45%). On the basis of these estimates, CH4 emissions from outdoor holding tanks for swine and dairy slurry in Canada were approximated at 0.71 Tg CH4 yr-1 (¿40%) and 0.24 Tg CH4 yr-1 (¿70%) respectively, giving a combined annual emission of approximately 0.95 Tg yr-1 (¿50%). ¿ 1998 American Geophysical Union