OBJECTIVES: This study investigated the correlation between the type of health insurance membership as a proxy for the economic status of patients and the severity of their type two diabetes mellitus (T2DM) in Indonesia. DESIGN: The study conducted a secondary analysis of National Health Insurance (Jaminan Kesehatan Nasional) claim data provided by the Indonesian Social Security Agency, Badan Penyelenggara Jaminan Sosial (BPJS). We used ordered logistic regression with four severity levels for T2DM (0=outpatient, I=mild, II=moderate, III=severe) as dependent variables. The main independent variables (insurance membership categories) included subsidised insurance members (PBI), a combination of formally employed and nonsalaried informal workers (PBPU & PPU) and nonworkers (BP). SETTING: Secondary healthcare facilities in Indonesia. PARTICIPANTS: The dataset included 2 989 618 claims for hospital visits of people with T2DM from 2018 to 2022. PRIMARY OUTCOME MEASURES: Severity level of T2DM patients. RESULT: A higher percentage of T2DM patients who visited healthcare facilities with subsidised insurance (PBI), which represents a low-income group, have severe disease (6.9%) than patients in the PBPU & PPU (4.9%) and BP categories (5.5%). Moreover, regression analysis revealed that having PBI membership status was associated with a greater OR of having severe T2DM than nonsubsidised members. Among T2DM patients in the nonsubsidised insurance category, workers (PBPU & PPU) had an OR of 0.74 (95% CI: 0.735 to 0.745; p<0.0001) for having severe disease during hospital visits. Moreover, non-workers (BP) had a lower OR of 0.718 (95% CI: 0.711 to 0.725; p<0.0001) for severe disease than the PBI category. CONCLUSION: These findings illustrate the lack of optimal access to health services for diabetes patients in low-income insurance membership categories and the challenges of better treatment in health facilities for low-income patients.