To achieve behaviour change, social advertisers have long utilised positive and negative advertising appeals employing different emotions and framing techniques. The evidence base indicates mixed results of effectiveness, making practitioner guidance on optimal advertising appeals unclear. Several gaps are evident in the literature that limit the understanding of maximising advertising appeal effectiveness to deliver behaviour change. First, the fragmented, mixed and inconsistent evidence within the literature of which emotional appeal is most effective limits practitioners understanding of how effective messages can be created. Systematic studies that synthesise knowledge can help to understanding which advertising appeals can be considered. Second, the literature is dominated with studies of fear and humour appeals, limiting understanding about the efficacy of other emotional appeals in creating positive behaviour change outcomes. This includes the lack of studies into emotional appeals utilising emotions such as shame and hope as well as co-active appeals where positive and negative emotions are utilised in an advertisement. Third, empirical examinations of the effectiveness of advertising appeals reported in peer review literature mainly employ indirect behavioural measures through self-reported data (e.g. attitudes and intentions), which do not always translate into behaviour. With the rapid evolvement of social media and machine learning methods many quantitative and qualitative metrics are readily available to evaluate advertising effectiveness more thoroughly. The use of such measures with the lack of validated frameworks for evaluating social media advertising messages is difficult for scholars and practitioners. Furthermore, current persuasion and social media marketing frameworks lack evaluation measures beyond social media engagements (e.g., likes) and intentions. This presents a need for a validated and reliable framework that takes the full suite of in social media engagement metrics and an more nuanced understanding of how to connect and track online behaviour change actions beyond social media to examine campaign effectiveness. Fourth, with the focus on conscious processes and measures in persuasion frameworks, unconscious measures remain underexplored. Scholarly calls are evident to investigate the role of sentiment on behaviours to further extend beyond self-reported data collection measures. Understanding the role of sentiment as an unconscious metric of public attitudes on behaviour enables governments and non-government organisations to focus attention on activities that reduce negative sentiment and increase public trust, helping in increasing public health campaigns’ effectiveness. In light of the aforementioned gaps, this thesis presents a series of published papers examining advertising appeal effectiveness with a core focus directed at understanding how advertising is linked to behaviour change.