Effects of ventriculectomy and prosthetic laryngoplasty on upper airway flow mechanics and blood gas tensions in exercising horses with induced left laryngeal hemiplegia were assessed. Five adult horses were trained to stand, trot (4.5 m/s), and gallop (7.2 m/s) on a treadmill (6.38 degrees incline). Inspiratory and expiratory airflows (VImax, VEmax, respectively) were measured using a 15.2-cm diameter pneumotachograph in a face mask. Inspiratory and expiratory transupper airway pressures (PuI, PuE, respectively) were determined as pressure differences between barometric pressure and lateral tracheal pressure. Blood collected from exteriorized carotid arteries was analyzed for PaO2, PaCO2, pH, hemoglobin (Hb) content, and HCO3- values. Heart rate (HR) was determined with an HR monitor. Measurements were made with horses standing, trotting, and galloping before left recurrent laryngeal neurectomy (LRLN; base line), 14 days after LRLN, 30 days after ventriculectomy (44 days after LRLN), and 14 days after prosthetic laryngoplasty (58 days after LRLN). Before LRLN (base line), increasing treadmill speed for horses from standing to the trot and gallop progressively increased HR, respiratory frequency, VImax, VEmax, PuI, PuE, Hb, and PaCO2 values and decreased PaO2, pH, and HCO3- values; inspiratory and expiratory impedances were unchanged. After LRLN, inspiratory impedance and PuI were significantly (P less than 0.05) increased in horses at the trot and gallop, and PaCO2 was significantly increased in horses at the gallop. The VImax and respiratory frequency were significantly (P less than 0.05) decreased in horses at the gallop. Left recurrent laryngeal neurectomy had no effect on PuE, VEmax, HR, PaO2, pH, Hb, or expiratory impedance values.(ABSTRACT TRUNCATED AT 250 WORDS)