Testing the equality of variances during hypothesis testing is an important preliminary step before using statistical tests such as the t-test or ANOVA. It has been demonstrated that many tests for equality of variances are sensitive to non-normal distributions. Using computer simulation, the present simulation study investigates the Type I error rate and statistical power of the nonparametric and median versions of the Levene test for equality of variances when there are three, four or five groups used in the analysis. For each of the three, four and five group conditions there are several levels of sample size, variance ratio, group sample size imbalance, and degree of skew in the population distribution included in the simulation. Results show that the nonparametric Levene test shows good statistical properties when samples come from heavily skewed population distributions, when overall sample size was small, and when groups were unbalanced. The findings also allow for a relative comparison of the median-based Levene test of equality of variances under a variety of conditions. Practical implications for the testing for equality of variances are discussed.