BACKGROUND Measurement of pH in exhaled breath condensate (EBC) is robust and simple. Acidic source fluid (airway lining fluid) traps bases while volatilising acids, leading to EBC acidification in many lung diseases. Lower airway ammonia is one determinant of airway lining fluid pH, raising the concern that addition of the base ammonia by contamination from the mouth might confound EBC pH assays. METHODS Three discrete methods were used to limit oral ammonia contamination of EBC collections: endotracheal intubation, oral rinsing, and -40 degrees C condenser temperatures. Separately, ammonia was removed from collected EBC samples by lyophilisation and resuspension. Intraweek and intraday variability of ammonia concentration was determined in 76 subjects, and ammonia and pH from a further 235 samples were graphically compared. Ammonia was assayed spectrophotometrically and pH was assessed after deaeration. RESULTS Data from 1091 samples are presented. Ammonia was reduced in EBC by all methods. Endotracheal intubation decreased EBC ammonia from a mean (SD) of 619 (124) microM to 80 (24) microM (p<0.001, n=32). Oral rinsing before collection also led to a decline in EBC ammonia from 573 (307) microM to 224 (80) microM (p=0.016, n=7). The colder the condensation temperature used, the less ammonia was trapped in the EBC. Lyophilisation removed 99.4 (1.9)% of ammonia. Most importantly, the pH of EBC never decreased after removal of ammonia by any of these methods. Intraweek and intraday coefficients of variation for ammonia were 64 (27)% and 60 (32)%, which is substantially more variable than EBC pH assays. CONCLUSIONS Although ammonia and pH appear to correlate in EBC, the oral ammonia concentration is not an important determinant of EBC pH. No precautions need to be taken to exclude oral ammonia when EBC pH is of interest. The low pH and low ammonia found in EBC from patients with lung diseases appear to be independent effects of volatile compounds arising from the airway.