Although airway disease in preschool children is common, standard spirometry is limited by the level of cooperation. We evaluated a computer-animated system (SpiroGame) aimed at improving children's performance in spirometry. SpiroGame includes a commercial pneumotachograph (ZAN100; ZAN Messgeraete GmbH, Oberthulba, Germany) and games teaching tidal breathing and all steps of an FVC maneuver. SpiroGame was compared with commercial flow-targeted candle-blowing software (MasterLab, Jaeger, Germany), and with extrapolated predicted values. Of 112 children aged 3 to 6 yr, 10 refused spirometry and 102 proceeded to FVC games and were randomized to initially perform either SpiroGame or candle-blowing. Training lasted 5 to 10 min for SpiroGame and 3 to 7 min for candle-blowing. Acceptable spirometry was performed by 69 of 102 children with SpiroGame and 48 of 102 with candle-blowing (p = 0.005). Order did not affect success. Acceptable FEV(1) maneuvers were achieved by 55 children with SpiroGame and two children with candle-blowing. The intrasubject coefficient of variation was 4.0% for FVC and 3.3% for FEV(1) with SpiroGame. A premature expiratory break occurred in 41 subjects with candle-blowing and in six with SpiroGame. FEV(0.5) could be measured with both systems. FVC and maximal midexpiratory flow at 50% of FVC (MMEF(50)) values were similar, whereas peak expiratory flow was higher with candle-blowing. In 39 healthy children, most parameters with SpiroGame were similar to extrapolated normal values. We conclude that an interactive computer-animated system facilitates successful spirometry in preschool children.