Tussometry is a new non-invasive technique for objectively assessing laryngeal function by analysis of the airflow waveform produced by a maximum effort voluntary cough manoeuvre. We describe the technique and present the calibration data. The tussometer has been calibrated for flows of up to 1100 litre min-1 using a flowmeter with a quoted accuracy of +/- 1.75%. The variables measured (cough peak flow rate (CPFR) and peak velocity time (PVT)) were found to be reproducible; the within-subject variability for CPFR was found to be 23.9% and for PVT 9%. There was no inter-observer variation. We found that the size of the mask used did not influence the readings obtained, providing an adequate seal was achieved.