The well-mixing assumptions, uniform distributions of air temperatures and contaminant concentrations and neglect of air momentum effects in a zone, used in multizone airflow network models could cause errors in some cases. Through dimensional analysis of the published data from the literature, this study has found correlation between the errors and mixing levels. Our study concluded that the assumption of uniform air temperatures is not appropriate if the vertical non-dimensional temperature gradient is higher than 0.03. The assumption of uniform air momentum effects could fail when the airflow with a strong momentum effect cannot be totally dissipated. The assumption of uniform contaminant concentrations can be invalid unless the contaminant source in the zone is also a heat source and the associated Archimedes number is greater than 400.