Hospitalisation and death from pertussis in highly immunised populations largely occurs before the first vaccination at 2 months. A Markov model was constructed to estimate the costs and health consequences of three strategies to reduce pertussis over the first 6 months of an infant's life. Earlier vaccination (at either birth or 1 month in addition to current practice) or vaccination of the parents soon after birth was compared with the current practice of vaccination at 2, 4 and 6 months. The model was populated using data on the incidence and costs from Australia. Disability-adjusted life-years (DALYs) were used as the primary outcome measure. The cost to the Australian public health system was chosen as the economic perspective, and Monte-Carlo simulations were used to accommodate uncertainties in the variables. Vaccination at birth was estimated to cost (S.D.) an additional A$33.21 (A$1.60) per infant and to reduce cases, deaths and DALYs by 45%. Vaccination at 1 month was estimated to cost an additional A$43.24 (A$8.98) per infant and to reduce morbidity by approximately 25%. Parental vaccination at birth was the most expensive alternative, costing an additional A$73.38 (A$4.98) per infant and reducing pertussis morbidity by 38%. The costs per DALY averted were A$330,175 (A$15,461) A$735,994 (A$147,679) and A$787,504 (A$48,075) for the birth, 1 month and parental vaccination strategies, respectively. Changing the estimated factor by which hospitalisations and deaths are under-reported, and the efficacy of early vaccination, had large effects on results. Parental vaccination at birth was most cost-effective where protection persisted for subsequent children. The birth vaccination strategy appears to offer the greatest potential benefit for one-child families, but the efficacy at birth (and 1 month) needs to be established.