Cooperation is defined as a ‘helping’ behaviour that provides direct fitness benefits to other individuals. Such behaviours have long intrigued evolutionary biologists, as it poses a problem for classic evolutionary theory, i.e. why should an individual perform a behaviour that is beneficial to other individuals? Indeed, an expansive body of work on evolutionary game theory, as well as, empirical studies, have since provided many mechanisms for promoting stable cooperation between unrelated individuals. Humans, however, often deviate from the optimal strategies predicted by theoretical models, which has emphasized the need to understand decision making processes. For example, the use of decision short cuts, known heuristics, allows individuals to make decisions quickly and accurately in frequently occurring situations, but may lead to less than optimal behaviour in novel contexts. Additionally, cognitive constraints, such as learning capabilities or failure to identify relevant environmental or social cues, may also cause deviations from predicated behaviour. Using blue streak cleaner wrasse (Labroides dimidiatus) as a model system, the primary aims of my PhD thesis were 1) to investigate important mismatches between standard theoretical predictions regarding animal decisions during cooperative interactions and experimental data, as well as, 2) to explore how well cleaners are able to readily identify and use relevant cues for decision making. Field work was conducted between 2011-2014, at Lizard Island Research Station, QLD, Australia.