Since most firms select December fiscal year-ends, theJanuary effect is a fiscal year-end accounting effect, according to the accounting-information hypothesis. This hypothesis attributes the unusually large stock returns in January to higher risk, caused by uncertainty about the impending announcement of firm performance. The empirical evidence does not support the hypothesis. Small firms with non-December fiscal year-ends fail to display a fiscal year-end effect. Yet all small firms, regardless of their fiscal year-end month, exhibit large January returns.