Abstract: Metadata are critical in epidemiological and public health research. However, a lack of biomedical metadata quality frameworks and limited awareness of the implications of poor quality metadata renders data analyses problematic. In this study, we created and evaluated a novel framework to assess metadata quality of epidemiological and public health research datasets. We performed a literature review and surveyed stakeholders to enhance our understanding of biomedical metadata quality assessment. The review identified 11 studies and nine quality dimensions; none of which were specifically aimed at biomedical metadata. 96 individuals completed the survey; of those who submitted data, most only assessed metadata quality sometimes, and eight did not at all. Our framework has four sections: a) general information; b) tools and technologies; c) usability; and d) management and curation. We evaluated the framework using three test cases and sought expert feedback. The framework can assess biomedical metadata quality systematically and robustly.