This paper analyzes influential factors of recommendation behaviour in social network sites (SNSs). Extant research on both SNSs and electronic word of mouth (eWOM) has given insufficient attention to SNSs as a potential eWOM channel. Considering the specificities of SNSs, this paper distinguishes implicit and explicit recommendation behaviour. Drawing upon research on eWOM, SNSs, and knowledge exchange, influential factors of implicit and explicit recommendation behaviour are identified. A theoretical model explaining why SNS users (do not) engage in implicit and explicit recommendation behaviour is developed. Structural equation modeling (SEM) is used for hypothesis testing. Data was collected via an online survey from 832 SNS users. The empirical results show a positive impact of reciprocity on both implicit and explicit recommendation behaviour, a negative impact of fear of producing spam on implicit recommendation behaviour, and a positive impact of both implicit recommendation behaviour and the perceived value of the recommended product on explicit recommendation behaviour.