The aim of this study was to determine the comparison of duplicate breath samples taken on a Drager Alcotest 7110, Mk II. A total of 216 persons were tested with the second sample being obtained 2 to 10 minutes after the first sample had been analysed. The mean difference between duplicate samples is 0.003 grams of alcohol in 100 millilitres of blood. The minimum difference was 0.000 grams of alcohol in 100 millilitres of blood and a maximum difference 0.015 grams. The Report of the Joint Legal, Policing and Scientific Committee determined that between 0.020 and 0.049 grams of alcohol in 100 millilitres of blood the second result must be within the criteria of ±0.010 grams of alcohol in 100 millilitres of blood of the first result. Above 0.050 grams of alcohol in 100 millilitres of blood the second result must be within the criteria of ±15 per cent of the first result. The failure rate between 0.020 and 0.049 (n=79) is zero with a standard deviation of 0.0014. The failure rate where the BAC is > 0.050 (n=101) is <1 per cent with a standard deviation of 0.0033. 36 subjects were less than 0.020 grams of alcohol in 100 millilitres of blood. Duplicate breath samples can be successfully obtained on a Drager Alcotest 7110 Mk II instrument coupled with a good quality assurance programme.