Streptococcus pneumoniae (pneumococcus) is one of the most important bacterial pathogens and a leading cause of mucosal infections (e.g. otitis media) and various forms of serious diseases (e.g. pneumonia, meningitis, bacteraemia) in developing and developed countries. Based on the polysaccharide capsule, there are at least 90 different pneumococcal serotypes, which may compete with each other to colonize the nasopharynx. Newly developed protein–polysaccharide conjugated vaccines have been shown to provide protection against disease caused by the serotypes included in the vaccine, and also against colonization (carriage). It is feared that yet uncommon, but nonetheless pathogenic serotypes which have been suppressed by competition, may become more prevalent in carriage and disease after large-scale use of conjugate vaccines. In this paper, we use transmission models of pneumococcal carriage to study how competition and vaccination influence the coexistence of two serotypes. According to our results, direct (physical) competition between two pneumococcal serotypes only influences colonization if the duration of naturally acquired immunity is short. By contrast, indirect (antibody-mediated) competition is of influence only if naturally acquired immunity is long lasting. Vaccination reduces the prevalence of the target serotype – an effect that is enforced by the presence of directly competing bacteria. The emergence of a non-target serotype after vaccination is only observed if bacteria compete directly. These results emphasize the importance of studying whether bacteria compete directly or indirectly and for how long people are protected in order to assess the long-term effects of sero-competition.