Measuring the efficiency of public higher education institutions has become a subject of many studies. We analyse these studies using meta-analysis and identify the most commonly adopted inputs and outputs in the DEA and SFA models (e.g. the number of students, graduates, and academic staff). Data obtained from these studies were used for meta-regression analysis. We analyse the effect of independent variables (sample size, the number of inputs and outputs, method used, model orientations, returns to scale and to the country) on the average technical efficiency of public higher education institutions. Finally, we use this model to predict the average technical efficiency of the Czech and Slovak public higher education institutions.