A nested PCR method was compared with culture for the detection of Bordetella pertussis in a routine clinical diagnostic laboratory. A total of 241 clinical nasopharyngeal aspirates were examined in parallel in the laboratory. Both methods were positive for 75 samples (31%), eight samples were positive by nested PCR only (3.3%), and one sample was positive by culture only (0.4%). The mean time actually required in the clinical laboratory (not operating with pertussis diagnosis during weekends) from the day of arrival to the diagnosis of a positive or negative sample by the nested PCR assay was 1.8 +/- 1.3 days (mean +/- SD), for positive culture 4.5 +/- 1.4 days and for negative culture 10.5 +/- 1.0 days. The hands-on time in the laboratory to perform the nested PCR was 2 h, for a positive culture 25 min, and for a negative culture 15 min. The cost analysis of the methods, when running one sample at a time, showed that the laboratory cost for PCR was six times higher than culture. When running four samples together the cost for PCR was three times higher than culture. In conclusion, the nested PCR is the more rapid and sensitive method compared to culture. With the present design, the PCR-protocol involves higher material expenditure and claims more hands-on time.