Purpose To investigate the role of complete blood cell count (CBC) measures in retinal artery occlusion (RAO). Methods This was a case-control study, including 73 newly diagnosed RAO patients and 73 sex- and age-matched subjects without RAO. On the same day of RAO diagnosis, a blood sample was collected and CBC was determined using an automatic blood counter. Dimensional CBC indices, such as mean platelet volume (MPV) and red cell distribution width (RDW), and some CBC-combined indices, including neutrophil/lymphocyte ratio (NLR), derived NLR [dNLR = neutrophils/(white blood cells - neutrophils)] and platelet/lymphocyte ratio (PLR), were evaluated. Erythrocyte sedimentation rate (ESR) was also measured. Results Median neutrophils, red cell distribution width (RDW), NLR and dNLR were 4.5x10(9)/L (IQR = 3.8-5.8), 13.4% (IQR = 12.7-14.75), 2.47 (IQR = 1.85-3.13) and 1.70 (IQR = 1.26-2.18) in RAO patients and 4x10(9)/L (IQR = 3.18-4.93), 12.9% (IQR = 12-14), 1.86 (IQR = 1.42-2.44) and 1.32 (IQR = 1.02-1.64) in controls. RAO patients had significantly higher values of neutrophils (p = 0.003), RDW (p = 0.0011), NLR (p = 0.0001) and dNLR (p = 0.0001). There were no significant differences between the values of white blood cells, lymphocytes, platelet count, MPV and PLR. Multivariate logistic regression models revealed a statistically significant correlation between RAO and increased RDW (OR = 1.36, 95% CI = 1.06-1.73, p = 0.015), NLR (OR = 2.02, 95% CI = 1.34-3.06, p = 0.0009) and dNLR (OR = 3.4, 95% CI = 1.71-6.75, p = 0.0005). Conclusion Results suggest that RDW, NLR and dNLR may be involved in the pathogenesis of RAO and predict its occurrence. However, high-quality epidemiologic studies, preferably of cohort design, are warranted to confirm whether, or not, an RDW, NLR and dNLR may be considered potential biomarkers of RAO.