When re-learning a motor skill, practicing a variety of treatment targets promotes error processing and the exploration of motor control strategies, which initially disrupts accuracy during training (motor performance), but ultimately enhances generalization, retention, and transfer (motor learning). Cough skill training (CST) is feasible and efficacious to improve cough strength; however, previous studies have used the same practice target during training. Our goal was to examine the impact of CST with variable practice on motor performance, motor learning, and respiratory system adaptations. The study was a prospective three-visit single group design. Twenty individuals with Parkinson's disease (PD) and concomitant dysphagia and dystussia completed two sessions of CST involving three randomized practice targets. Cough, lung volume, and airway clearance outcomes were assessed before and after treatment sessions with long-term retention evaluated after 1 month. Peak expiratory flow rate improved after CST with variable practice for voluntary single (β = 0.35 L/s) and sequential (β = 0.22 L/s) cough, which were maintained after 1 month without treatment. The ability to expel material from the upper airway demonstrated a small magnitude of improvement (β = -1.87%). During CST, participants altered lung volume based on the treatment target and lung volume decreased during reflex cough after completing CST. Individuals with PD demonstrated improvements in several aspects of motor learning after two sessions of CST with variable practice. Increasing lung volume may not be an implicit strategy to upregulate voluntary cough strength in this treatment paradigm. The findings support the need for larger investigations exploring the potential benefits of this CST approach.