A unique data set from the Tel-Aviv Stock Exchange (TASE) is used to study the effect of trading mechanisms on stock return volatility. The TASE represents a natural experiment which allows separation of the overnight information effect from the trading mechanism effect. The data span a time period in which the order of the trading mechanisms (a sequential continuous mechanism and a call auction) was switched. Since overnight information should impact opening prices equally across periods, this affords an unparalleled opportunity to examine the trading mechanism effect without the confounding effect of the non-trading period. This paper finds that the null, that opening variances equal closing variances, cannot be rejected for either period. Further, the tests cannot reject null that the ratio of opening to closing return variances is equal across periods. This suggests that the trading mechanisms on the TASE do not differ in their effect upon return volatility.