The aim of the paper is to estimate the day of the week effect in the stock markets in the Czech Republic, Hungary and Poland over the period 2006 – 2012. The entire period of estimation is divided to six sub-periods capturing individual phases of the financial and economic crisis. We separately estimate a modified GARCH-M (1,1) model for each country and each sub-period using daily returns of the major national stock market indices. The day of the week effect is measured for both daily returns and conditional variance (volatility) of the returns. The results clearly indicate that there is a little evidence of day of the week effect. Daily calendar anomalies are rather sporadic, isolated, unstable over time and often opposite to theoretical assumptions. There is no phase of financial crisis characteristic of significantly increased incidence of day of the week effects. We conclude that the day of the week effect is not typical for the Central European stock markets and the recent financial crisis seems to have no impact on existence of this phenomenon in the markets.