The problem of estimating suspended sediment yields from ungauged watersheds is described. Results from regression analyses using hydroclimatic and geomorphic variables are presented for 14 watersheds in Atlantic Canada. Estimates are given for long-term mean suspended sediment yields and for annual suspended sediment yields for a particular year. The effects of incorporating an added constant term in the regression equation are considered. Limited tests are presented for two small watersheds not used in the development of the adopted regression equations.Key words: suspended sediment yields, ungauged watersheds, regression analysis, Atlantic Canada.