Purpose <p>This study explored the correlation between Hemoglobin A1c (HbA1c) and intraocular inflammatory cytokines in proliferative diabetic retinopathy (PDR) patients, aiming to provide evidence for PDR prevention and management.</p> Methods <p>This was a cross-sectional study involving 100 patients and 100 eyes diagnosed with PDR. All patients were collected demographic characteristics, HbA1c, and history of chronic diseases at baseline. Aqueous humor samples of patients were obtained during the intraocular surgery. Then 10 cytokines were assessed using the Cytometric Beads Array (CBA) technique. The relationships between HbA1c and each inflammatory factor were further tested by restricted cubic splines (RCS) in the logistic model.</p> Results <p>A U-shaped curve was found between IL-6 (interleukin-6) and HbA1c (<i>p</i> = 0.0001). HbA1c was negatively associated with IL-6 when below 8.31%. There was also a U-shaped exposure-response relationship between vascular endothelial growth factor (VEGF) and HbA1c (<i>p</i> = 0.0211), with the lowest VEGF levels observed at an HbA1c of 6.72%. A J-shaped curve was observed in the relationship between IL-8 (interleukin-8) and HbA1c (<i>p</i> = 0.0061). When HbA1c exceeded 8.23%, IL-8 exhibited a significant upward trend. Monocyte chemoattractant protein-1 (MCP-1) shows a positive linear relationship with HbA1c (<i>p</i> = 0.0036). No significant differences were found between other cytokines and HbA1c.</p> Conclusions <p>Our study identified the correlation between HbA1c and intraocular inflammation in PDR, characterized by U-shaped relationships with IL-6 and VEGF, a J-shaped relationship with IL-8, and a linear relationship with MCP-1. These findings advocated for a nuanced approach to glycemic management. However, longitudinal studies are still needed to determine whether HbA1c fluctuations directly drive cytokine changes.</p> Clinical trial registration number <p>Not applicable.</p>