Background: Bronchoalveolar lavage (BAL) cytology results from 1 lung might not be representative of both lungs. Objectives: To determine whether the lung site sampled would influence the horse's BAL cytology profile, and if a pooled BAL sample would be superior with regard to BAL cytology diagnosis in a cohort of healthy and subclinical asthmatic warmblood horses. Animals: Fifty-nine horses in 2021 and 70 horses in 2022, the follow-up included 53 of the same in each year. Methods: A cross-sectional study with follow-up included BAL cytology samples from individual lungs and from pooled BAL samples. The BAL samples were enumerated and differential cell count were applied to categorize the horses as control or with airway inflammation (AI). Results: Bronchoalveolar lavage mast cell count was higher in left lung compared to right lung (2021; median 1.6 [range, 0.6-3.3] vs 1.2 [0.7-1.5] P =.009, 2022; median 3.1 [2.1-4.2] vs 2.4 [1.7-3.4], P <.001) and compared to pooled samples (2022; median 2.6 [1.7-3.7], P <.001). Between year 2021 and 2022, 17 of the horses had changes in BAL cytology from control to AI or vice versa. Conclusions and Clinical Importance: Pooled BAL sample was the least reliable for detecting AI, and was not representative of the overall lung condition.