Purpose: This meta-analytical study aims to test the interrelationships within Bolman and Deal's four-frame leadership model in educational settings around the world. Moreover, we evaluated the effect of several factors like publication year, study type, sample group, level of education and country of data collection on these relationships. Design/methodology/approach: Data covered the articles and theses between 1994 and 2022, and the dataset consisted of 23 independent studies with an identical number of groups and a total sample of 5,835. The analyses proved that there was no publication bias. Fisher [Z with swash tail] Coefficient was reported for the average effect size analyses in line with the random effects model. Findings: We concluded that the average effect size of the relationships among all four frames was large. Subgroup analyses yielded significant differences by study type, sample group, level of education and country of data collection, but not by publication year. Research limitations/implications: While the study sample comprises a well-known leadership model, it excludes some other popular leadership models (e.g. instructional leadership and transformational leadership) in the field. Nevertheless, research findings elaborate on the complex nature of school leadership. Originality/value: The results indicate the intercorrelated nature of Bolman and Deal's four-frame leadership model in support of the proposition for using multiple frameworks in combination. We also concluded that this effect is beyond cultural differences, antecedents and consequences.