Abstract Implicit taxes reflect the extent (if any) to which tax-favored assets bear lower pretax returns than do tax-disfavored assets of similar risk. Prior research on implicit taxes has met with mixed results, particularly in equity securities, because of the difficulty in separating tax effects from effects caused by cross-sectional differences in risk. We avoid problems of risk by essentially comparing each security to itself before and after an unexpected change in the manner in which dividends are taxed to corporate investors. We find strong evidence of implicit taxes in preferred stocks. Extensive testing using the same event date indicates that no similar implicit tax effect exists in common stocks.