The objective of this paper was to determine if changes in ventilation patterns could influence the outcome of respiratory function measurements performed with our impulse oscillometry system (IOS) in horses. In a first study, IOS tests were performed in vitro on six isolated equine lungs. Lung inflation levels were controlled by modifying depressurisation inside an artificial thorax and different ventilation patterns were imposed. In a second in vivo study, transient variations in breathing pattern were evaluated both with the IOS and a current reference technique (CRT) in five healthy mature horses after an intravenous (i.v.) injection of lobeline hydrochloride. In both studies, respiratory rate (RR, range: 7-42 breaths/min.) and tidal volume (V(T), range: 0.4-25 L) had minor or no influence on IOS parameters. The influence of lung inflation, most marked for resistance at 5 Hz (R(5 Hz)), was limited for the considered physiological range. In vivo, statistical models indicated that maximal changes in pleural pressure (Max Delta Ppl) and peak flows were the main determinants of the variability of the resistance (R(rs)) and the reactance (X(rs)) of the respiratory system. The fourfold increase in baseline Max Delta Ppl and peak flows obtained during hyperpnoea caused a significant increase in R(rs) at 5 and 10 Hz and a decrease in X(rs) at all frequencies. We conclude that IOS parameters are not influenced by tachypnoea, but will reflect alterations in respiratory mechanics caused by hyperpnoeic breathing.