AimsThis study aims to investigate the relationship between long‐term visit‐to‐visit within‐person HbA1c variability and hospitalisation outcomes in adults with type 2 diabetes (T2D).MethodsWe conducted a cohort study at a tertiary hospital in Singapore involving people aged 21 to 101 years with T2D who had ≥3 HbA1c tests over 2 years. HbA1c variability was assessed using coefficient of variation (CV), variability independent of the mean (VIM) and HbA1c variability score (HVS). A 1‐year follow‐up was performed after the last HbA1c measurement to identify all‐cause and potentially avoidable hospitalisations (PAH), categorised as overall, acute, chronic and diabetes composites.ResultsThe study included 14 923 patients (mean age: 62.9 ± 12.9 years; 55% male). The median HbA1c variability was 8.6% CV (IQR: 5.1–14.3). Higher quartiles of HbA1c variability were associated with greater risks of PAH and all‐cause hospitalisations, independent of glycaemic control. Compared to Q1, for example, the risk ratios and 95% confidence intervals for diabetes‐related PAH based on HbA1c CV were as follows: Q2, 1.32 (0.93–1.88); Q3, 1.65 (1.18–2.31) and Q4, 2.16 (1.54–3.03). For all‐cause hospitalisations, they were as follows: Q2, 0.97 (0.90–1.05); Q3, 1.08 (1.00–1.17) and Q4, 1.16 (1.07–1.26). When stratified by glycaemic control, elevated risk of PAH persisted even in those with optimal glycaemic control. Consistent findings were observed using HbA1c VIM and HVS measures.ConclusionsIn individuals receiving care at specialist outpatient clinics of a tertiary hospital, HbA1c variability is associated with a higher risk of PAH. Comprehensive diabetes management strategies addressing both glycaemic control and variability may offer benefits.