This prospective study, which was designed to compare nasopharyngeal culture, polymerase chain. reaction (PCR) and serology in the diagnosis of pertussis, covered 35 children aged between 0 and 16 who were admitted to Hacettepe University Ihsan Dogramaci Children's Hospital between 1 March 2005 and 31 August 2006 with coughing for 7 days or longer, paroxysmal cough of any duration, or cough with inspiratory whoop and/or vomiting (or apnea) after coughs. The demographic data and vaccination history of the patients were recorded. During the initial examination, samples were taken from the posterior nasopharynx for Bordetella pertussis (B. pertussis) culture and PCR analysis. In order to determine antibody positivity and antibody levels against B. pertussis antigens, serum samples were taken during the initial examination (acute phase) and two weeks later (convalescent phase). In the first serum sample, immunoglobulin M (IgM) was determined against pertussis toxin. In the first and second samples, IgA and IgG antibodies were evaluated against pertussis toxin and filamentous hemagglutinin. Culture yielded negative results in all of the patients. PCR was positive in two cases (5.7%). In the PCR-positive patients, IgM, IgA and IgG type anti-pertussis antibodies were found to be positive in the first serum samples; and IgA and IgG antibodies were found to be positive in the second serum samples. Therefore, it was considered that serology could be as sensitive as PCR when type IgM, IgA and IgG antibodies were found to be positive against a minimum of two antigens of B. pertussis. In conclusion, both PCR and serologic tests -if evaluating all types of antibodies to a minimum of two antigens of B. pertussis obtained in both acute and convalescent sera- could be more sensitive than culture in the diagnosis of pertussis.