This study tests the tax-induced trading hypothesis as an explanation of the turn-of-the-year anomaly using Canadian and U.S. intraday data. Since the Canadian tax year-end precedes the calendar year-end by five business days, tax effects may be isolated. We find the anomaly is related to the degree of seller-and buyer-initiated trading and depends upon the incidence of the taxation year-end. Seller-initiated transactions (at bid prices) dominate until the tax year-end after which buyer-initiated trades (at ask prices) dominate. The anomaly is a function of bid-ask prices.