Non-invasive techniques for monitoring ventilation in infants are widely used in short-term laboratory-studies but have not been evaluated in routine clinical settings. To determine whether respiratory inductance plethysmography (RIP) can provide reproducible measurements of tidal volume (VT) in premature infants over an extended period of time, we monitored respiration in eight healthy preterm infants over 4.9 +/- 1.0 hours (mean +/- SD). The algebraic sum (Sum) of rib cage (RC) and abdominal (AB) motion signals (obtained by RIP) was calculated and presented over the entire recording period as percent of an initial 5 minute calibration period. VT was simultaneously measured with a nasal mask pneumotachometer with infants in prone and supine positions during active and quiet sleep. Infants were studied in the morning (AM) and again in the afternoon (PM). Between these studies they were returned to the nursery wearing the RIP in a continuous record mode. For all patients there was a significant linear relationship between VT (in mL measured by pneumotachometer) and Sum (in % of calibration value, RIP). Neither the slope of the relationship (0.074 +/- 0.03 in AM vs. 0.071 +/- 0.02 in PM), nor its variability as measured by standard error of the estimate (SEE) (2.3 +/- 0.5 in AM vs. 2.5 +/- 0.8 in PM) changed significantly from AM to PM. The relationship between VT and Sum, as well as the variability of that relationship, was not altered by position, asynchrony of RC and AB, respiratory rate, or percent RC contribution to Sum. We conclude that RIP produces consistent measurements of respiratory effort over 5 hours in healthy preterm infants without need for recalibration and is not affected by routine care.