Empirical evidence on how health literacy affects diabetes outcomes is inconsistent. The purpose of this meta-analysis was to quantitatively summarize the findings on the associations between health literacy and diabetes knowledge, self-care activities, and glycemic control as disease-related outcomes, with specific focus on the type of health literacy assessment. Nine databases (MEDLINE, CINAHL, Communication and Mass Media Complete, PsychInfo, PsychArticles, Psychology and Behavioral Sciences Collection, ERIC, Sociology, Embase) were searched for peer-reviewed original research articles published until 31 March 2018. Studies with type 1 and/or type 2 diabetes patients aged 18 or older, providing a calculable baseline effect size for functional health literacy and diabetes knowledge, self-care activities, or HbA1C were included. The meta-analysis includes 61 studies with a total of 18,905 patients. The majority were conducted in the USA, on type 2 diabetes patients, and used the S-TOFHLA as a performance-based or the BHLS as a perception-based measure of functional health literacy. Meta-analytic results show that all three outcomes are related to health literacy. Diabetes knowledge was best predicted by performance-based health literacy measures, self-care by self-report measures, and glycemic control equally by both types of health literacy assessment. Health literacy plays a substantial role in diabetes knowledge. Findings for the role of health literacy in self-care and glycemic control remain heterogeneous, partly due to the type of health literacy assessment (performance- vs. perception-based). This has implications for the use of health literacy measures in clinical settings and original research. This meta-analysis was limited to functional health literacy and, due to the paucity of studies, did not investigate the role of other dimensions including communicative and critical health literacy.