The impact of different sampling techniques on the results of breath analysis was to be assessed in this study. Alveolar, mixed expiratory and time-controlled samples were collected from ten volunteers and from eight lung cancer patients. Breath sampling was visually controlled by means of capnometry. PCO(2) and 13 VOCs were determined. Mixed expiratory sampling yielded 25% lower concentrations of CO(2) and blood-borne VOCs. Time-controlled sampling generated high variation of results. Ratios C(alv)/C(mixed) were >1.5 for CO(2), acetone and isoprene, and <1 for isopropanol, 2-butanone and hexanal. Acetonitrile, butane, dimethylsulfide, pentane, butanal, benzene and hexane showed 1.5 > C(alv)/C(mixed) > 1. The ratio C(alv)/C(mixed) of CO(2), acetone and isoprene was different in healthy volunteers and lung cancer patients. Alveolar samples showed the highest concentrations of endogenous and lowest concentration of exogenous substances. Sampling can impact results in breath analysis. Valuable information can be obtained from ratios of alveolar and mixed expired concentrations.