Apparent central apnoea (absent breathing movements) detected by monitoring movement of the thoracic wall was compared with simultaneous detection by abdominal wall movement. Eighteen infants provided one or more 24 hour recording of heart rate (electrocardiography), thoracic respiration (transthoracic impedance), and abdominal wall movement (pressure sensitive capsule distortion). Detection of true apnoea, recognition of artefact, and measurement of the duration of true apnoea were all improved when two channels of respiratory monitoring were used in combination. We recommend that any study purporting to observe breathing patterns by indirect recording of respiratory movement will be more reliable if more than one channel of respiratory movements is recorded simultaneously. Further, in infants no estimation of duration of central apnoea can be made on the basis of either a transthoracic impedance record alone or an abdominal wall movement sensor alone. Comparison of findings among studies using different single channel recordings are unlikely to be meaningful.