Many patients in pre-hospital and emergency care are under the influence of alcohol. In addition, some of the more common pathological conditions can introduce a behaviour that can be mistaken to be related to alcohol inebriation. Fast quantitative determination of the breath alcohol concentration (BrAC) in emergency patients facilitates triage and medical assessment, but shallow expirations performed by non-cooperative patients reduce the measurement reliability. The aim of this study was to evaluate if breath alcohol analysis in non-cooperative patients can be improved with use of simultaneous measurement of the expired carbon dioxide (CO2). With prototypes of a handheld breath alcohol analyser based on infrared transmission spectroscopy the alcohol and CO2 concentration in expired breath from 37 cooperative and non-cooperative patients were measured. The results show that enhanced breath sampling with use of a pump and estimation of the end expiratory BrAC with use of the ratio between the measured partial pressure of CO2 ($$ P_{{{\text{CO}}_{2} }} $$) and a reference value of the alveolar $$ P_{{{\text{CO}}_{2} }} $$, provided adequate correlation with the blood alcohol concentration (BAC). This pre-clinical study has shown that breath alcohol analysis in shallow expirations from non-cooperative patients can be improved with use of CO2 as a tracer gas.