The burden of pneumococcal carriage is largest in developing countries from which, however, detailed studies on pneumococcal transmission are missing. In this study we followed nasopharyngeal carriage in Bangladeshi infants (n=99) from birth, with 2-week sampling intervals until age 4 months, and monthly thereafter until age 1 year, and also their family members at the same intervals. We assessed the dependence of pneumococcal acquisition rates on age, serotype, serotype-specific exposure (i.e. transmission) and current state of carriage (yes/no). A statistical model of pneumococcal transmission, taking into account incompletely observed data, was applied to estimate rates of acquisition and clearance for a large number of serotypes at the same time. Serotypes that were common in the study population were more often acquired from the community than rarer serotypes. However, when conditioning on serotype-specific exposure within the family, transmission rates were similar between different serotypes. Exposure within families signified more than tenfold increase in the rate of acquisition.