Background: Cytology of airway samples is sensitive for diagnosis of exercise‐induced pulmonary haemorrhage (EIPH), but the association between tracheal wash (TW) and bronchoalveolar lavage fluid (BALF) is unknown. The objective of this study was to determine whether diagnosis of EIPH, using haemosiderophages/macrophages (H/M) ratio, differs when based on TW or BALF. Methods: A prospective cross‐sectional study was conducted on 102 standardbred horses in training. TW and BALF were collected concomitantly from all horses at rest (at least 24 hours after their last training or race), and their H/M ratios were calculated. Spearman's correlation, Cohen's kappa and Gwet's coefficient tests were performed to evaluate the association between TW and BALF samples. Results: With BALF, 21 horses met the cytological inclusion criteria for an EIPH diagnosis from individual and/or pooled samples. With TW, 20 horses had occasional (H/M < 10%) haemosiderophages, and nine, one and three horses had small (10%–25%), moderate (25%–50%) and large (>50%) proportions, respectively. Poor correlations and inconsistent concordances between TW and BALF were found for H/M ratio. Limitations Limitations include the use of a single staining method and the absence of a total haemosiderin score. Conclusion: No association between TW and BALF was found for the cytological diagnosis of EIPH. Based on H/M ratio, BALF remains the sample type of choice for cytological diagnosis of EIPH.