I examine return seasonality in the foreign exchange market using currency futures during the period 1973−2015. All the G10 currency futures yield negative returns in January and this effect happens more often in the countries that have a tax year ending in December. In contrast, returns offered in April are positive. To exploit these anomalies, I use a seasonality strategy that selects portfolios based on their historical same-calendar-month returns. I find that this strategy does not work in the currency market, although I find consistent results with Keloharju et al. in the stock portfolios.