Introduction: In 2012 a large epidemic of pertussis occurred in the Netherlands. We assessed pertussis toxin (PT) antibody levels in longitudinal serum samples from Dutch 10-18 year-olds, encompassing the epidemic, to investigate pertussis infection incidence. Methods: Blood was sampled in October 2011 (n = 239 adolescents), then 1 year (2012; n = 228) and 3 years (2014; n = 167) later. PT-IgG concentrations were measured by immunoassay and concentrations ≥50 IU/mL (seropositive) assumed indicative of an infection within the preceding year. Results: During the 2012 epidemic, 10% of participants became seropositive, while this was just 3% after the epidemic. The pertussis acquisition rate proved to be sixfold higher during the epidemic (97 per 1,000 person-years) compared with 2012-2014 (16 per 1,000 person-years). In 2012, pertussis notifications among adolescents nationwide were 228/100,000 (0.23%), which is at least 40 times lower than the seropositivity percentage. Remarkably, 17 of the 22 seropositive participants in 2011, were still seropositive in 2012 and nine remained seropositive for at least 3 years. Discussion: Longitudinal studies allow a better estimation of pertussis infections in the population. A PT-IgG concentration ≥50 IU/mL as indication of recent infection may overestimate these numbers in cross-sectional serosurveillance and should be used carefully.