Summary Background Cytology plays a major role in the diagnosis of ocular surface diseases. Objective To compare 2 cytological sampling methods for obtaining corneal and conjunctival cell samples regarding irritation for the patient, invasiveness, sample quality and diagnostic equivalence. Study design Observational prospective study. Methods In 5 healthy horses, conjunctival and corneal samples were taken bilaterally by impression cytology sampling (ICS) and cytobrush sampling (CBS). Irritation and invasiveness were assessed with an eye irritation and an epithelial damaging score system, respectively. Sample quality was evaluated via morphometric analysis and graded by a board certified clinical pathologist. For the assessment of diagnostic equivalence, 15 eyes of 14 client owned horses with ocular surface anomalies were sampled by ICS and CBS. The methods were compared regarding the types of inflammatory cells and/or infectious agents detected and if the correct diagnosis could be achieved. Histopathology served as gold standard when available. Results ICS was significantly less invasive and less irritating for the horses. Both methods retrieved cells of overall high quality; the cell quantity was significantly higher in IC samples. ICS preserved the natural cellular layout. There was a fair, but no statistically significant agreement between the diagnostic outcomes between sampling methods although CBS resulted in a slightly greater variability of inflammatory cell types compared to ICS. Main limitations None. Conclusions Because of its low irritability and invasiveness ICS can be recommended for cell sampling in fragile corneas and for experimental studies. ICS is especially useful in cases where preservation of cellular layout is advantageous. CBS samples are easier to obtain because of the small equine palpebral fissure. Additionally, the identification of inflammatory cells within conjunctival cell samples is easier. Overall, CBS is still an appropriate method in clinical settings.