Variable obstruction to airflow at the laryngeal level may cause respiratory distress during exercise. The Continuous Laryngoscopy Exercise (CLE)-test enables direct visualization of the larynx during ongoing exercise. The aims of this study were to establish a scoring system for laryngeal obstruction as visualized during the CLE-test as well as to assess reliability and validity of this scoring system. Continuous video recording of the larynx was performed in parallel with continuous video recording of the upper part of the body, and recording of breath sounds in 80 patients and 20 symptom-negative volunteers, running on a treadmill to respiratory maximal tolerable distress or exhaustion. Each participant scored the degree of symptoms during exercise. The scoring system contains four sub-scores, each graded from 0 to 3. Two independent laryngologists, blinded to clinical data, scored the video recordings of the larynx twice. The proportion of inter- and intra-observer agreement (equal scores) for each sub-score through these four sessions varied between 70 and 100% (weighted kappa values varied from 0.49 to 1.00 correspondingly). A positive correlation was found between CLE-test sum score and symptom score (rho = 0.75, P < 0.001). There was a significant difference in CLE-test sum score between patients (3.34 +/- 1.34) and volunteers (0.65 +/- 0.66) (P < 0.001). The single CLE-test sub-score that correlated most strongly with symptom score was glottic adduction at maximal effort (rho = 0.75, P < 0.001). The presented scoring system is reliable and valid, and we suggest that it can be used when laryngeal function during exercise is evaluated.