There is no current consensus among published guidelines on whether noseclips are required during spirometry testing. This study investigated the effect of noseclips on spirometric measurements in patients with a range of disease. Fifty-two patients (30 male; mean age 58.0 years, range 19-78; mean FEV1 82.6% predicted, range 23.8-128.3%) performed measurements of VC, FVC and FEV1 according to ARTP/BTS guidelines (1994) using a wedge bellows spirometer (Vitalograph Model S, Bucks, U.K.). All patients performed two sets of measurements (with and without noseclips) in random order (Group 1 = noseclips first; n=30; Group 2 (without noseclips first, n=22). Tests were conducted by qualified physiologists. Measurements obtained with and without the use of noseclips were similar (mean differences FEV1 -0.030 L SD 0.210 and -0.005 L SD 0.093 for Groups 1 and 2 respectively; FVC -0.007 L SD 0.109 and -0.040 L SD 0.117; VC 0.036L SD 0.137 and -0.040 L SD 0.150) and were not dependent on patient group or previous test experience. Four patients had differences outside the 95% confidence limits for each parameter. There were no significant correlations between the differences with and without noseclips and severity of lung disease, age, smoking history, BMI or lung volume (all P > 0.100). The within patient coefficient of variation did not depend on the testing method. Use of noseclips during spirometry does not systematically affect the results obtained or the within-subject repeatability. Marked individual differences highlight the importance of maintaining consistency in the method adopted for a particular patient.