Recent studies have utilized nasal lavage to study the inflammatory cells of the nasal epithelium. In unexposed subjects, investigators have reported wide interindividual variability in lavage cell counts. The intraindividual variability of cell counts in sequential lavages has been less well described. Investigators have also reported that nasal lavage may washout cells, resulting in lower cell counts on subsequent lavages. The present study was designed to characterize both the variability in cell counts in unexposed volunteers and the kinetics of cell washout. Twenty-one subjects participated in two nasal lavage trials. In Trial 1, a baseline lavage was followed by a lavage 72 h later; in Trial 2, the baseline lavage was followed by a lavage 48 h later. Intraclass correlation coefficients of reliability (R) were calculated for each trial. In Trial 1, the R was 0.88, with a one-sided confidence interval > or = 0.75, whereas in Trial 2 R was 0.67, with a confidence interval > or = 0.40. The smaller R in Trial 2 may suggest that washout was more evident at 48 h than at 72 h after the baseline lavage. Furthermore, these R values suggest that within-subject variability is smaller than between-subject variability, supporting the utility of nasal lavage as a reliable technique for investigating the nasal cavity response to air pollutants.