The performance of two mixing height models are compared to observed mixing heights during convective conditions. The first model integrates the hourly surface heat flux and friction velocity to compute convective and mechanical estimates, respectively. The second model uses an interpolation approach that is currently recommended by the U.S. Environmental Protection Agency for use in air quality regulating analysis. The results suggest that both models perform best when estimating late afternoon mixing heights; nearly 85% of the estimates from both models are within 40% of the observed values. For the entire daytime period, the first model produced the least biased results; all the estimates, on average were within a factor of two of the observed mixing heights. The second model performed poorly in the hours shortly after sunrise, but was comparable to the first model for the latter part of the day.