<p>Slot-machine gambling is disproportionately associated with problem gambling. It is therefore important to develop an understanding of how the programming and features of slot machines influence gambling. The current research programme investigated a major feature of many slot machines which has so far been neglected by experimental research – the free-spins bonus feature. The first major focus of this research was a series of experiments that investigated whether participants prefer to play a slot-machine simulation with a free-spins bonus feature. In each experiment participants gambled on two simulations, one with a free-spins feature and a similar machine without free spins. Following this, participants could switch between the same simulations and the number of spins they made on each simulation was measured. Participants preferred the free-spins simulation, but only when the free-spins feature incorporated an increased rate of wins, music, animations, and graphics advertising the presence of the free-spins feature. In the final experiment investigating preference, we investigated whether responding was influenced by whether participants gambled with hypothetical money, or credit that could be exchanged for tangible rewards. Participants preferred the complex free-spins simulation similarly regardless of what they were gambling with, but selected larger bet amounts and gambled for longer when gambling with hypothetical money. The second major focus of this research programme was an investigation of whether free-spins features cause increased gambling persistence – a hallmark of disordered gambling. We developed a new persistence-measuring task which was adapted from research investigating Behavioural Momentum Theory. Participants gambled on two simulations in a multiple schedule design. An initial baseline phase consisted of four alternations of each component, one of which had the complex free-spins feature demonstrated to increase preference in the earlier experiments. Baseline phases then alternated with disruption phases where video-clips were embedded into the top right corner of the simulations. The rate at which participants gambled during baseline was compared to the rate at which they gambled when the videos were present, with bigger relative decreases in response rate during the disruption phases indicating less persistence. The free-spins feature did not cause participants to gamble more persistently. Further experiments also assessed whether different frequencies of wins caused participants to gamble more persistently, and results indicated that they did not. The findings of this research programme have implications for Behavioural Momentum Theory, suggesting that comparing response rate during disruption to response rate during baseline is not highly sensitive to small differences in reinforcement schedules. The findings are also relevant for slot-machine legislation, providing an indication that limiting or removing free-spins features may reduce player enjoyment without reducing persistent gambling. Furthermore, the task developed in the persistence investigation provides a useful tool which can be used to investigate how other features of slot machines influence persistence. Future research could, for example, investigate how free-spins features interact with other slot-machine features to influence gambling behaviour.</p>