The recent increase in Bordetella pertussis incidence (whooping cough) presents a challenge to global health. Recent studies have called into question the effectiveness of acellular B. pertussis vaccination in reducing transmission. Here we examine the epidemiological consequences of an ineffective B. pertussis vaccine. Using a dynamic transmission model, we find that: 1) an ineffective vaccine can account for the observed increase in B. pertussis incidence; 2) asymptomatic infections can bias surveillance and upset situational awareness of B. pertussis; and 3) vaccinating individuals in close contact with infants too young to receive vaccine (so called "cocooning" unvaccinated children) may be ineffective. Our results have important implications for B. pertussis vaccination policy and paint a complicated picture for achieving herd immunity and possible B. pertussis eradication.