The main goal of this thesis was to examine the influence of face and non-face emoji as a means to increase processing fluency and ratings of rapport in positive and negative contexts. This thesis aimed to draw on two key theoretical frameworks – the Processing Fluency Framework and the Rapport Management Model. In an initial naturalistic analysis (Chapter Two), the prevalence and types of emoji used were investigated, along with their relationship to self-presentation and related variables. Face but not non-face emoji were found to be linked to self-presentation variables, although the effects were weak. These emoji informed the design of subsequent experiments. In a series of five experiments (Chapters Three to Seven), the effect of face and non-face emoji on processing fluency and rapport were examined across positive and negative message contexts and manipulating a series of variables of relevance to the emoji (e.g., type, position, congruency with message). In each experiment, participants were presented with hypothetical text messages between friends and asked to rate them on a series of measures relating to fluency (efficiency, clarity, and/ or understandability and believability) and rapport (interest in the friendship and improving the friendship). Consistent with previous literature, emoji presence affected processing fluency and rapport. However, the effect varied depending on message valence, emoji types and the specific message content. Overall, the findings suggest a connection between processing fluency and rapport, related to the perception of emoji in text messages, a relationship which to date has not been identified in the literature. The findings, while supporting the processing fluency account, suggest that emoji effects are more complex, context dependent and nuanced than originally expected.