To describe the use of glycated haemoglobin (HbA1c) and fasting plasma glucose results in guiding treatment changes in patients with type 2 diabetes mellitus.Setting: Public sector primary health care facilities in the Cape Town Metropolitan Region in South Africa.Method: Retrospective, descriptive study design was employed. Data for an 18-month period were collected during 2014 and 2015. Data were collected from patient medical records and included baseline demographics, laboratory monitoring tests and the patients' last three prescriptions. Results:The study consisted of 575 participants (64% female) with an average age of 57 (± 11.38) years. The average baseline HbA1c for 493 participants with at least one result was 8.78% (± 1.63), and only 28% of these participants reached the glycaemic target at consequent consultations. HbA1c levels were available to guide 245 prescription changes, of which 181 of these results were outside of the target range. Of these, 15.5% had appropriate therapy adjustments, 78.4% had no change or a lateral change in their follow-up prescriptions, and 6.1% had therapy adjustments opposite to what guidelines suggest. Conclusion:Glycaemic monitoring indicated consistent suboptimal glycaemic control in more than 60% of participants. Medicine prescribing patterns did not align with the prescribed local guidelines, Society for Metabolism, Endocrinology and Diabetes of South Africa (SEMDSA). The appropriate use and interpretation of HbA1c at a clinic level should be emphasised to promote rational use of medicines that minimise acute hospitalisation episodes and optimise patients' long-term health outcomes.