Recently, swimming-style colour synaesthesia was introduced as a new form of synaesthesia. A synaesthetic Stroop test was used to establish its genuineness. Since Stroop interference can occur for any type of overlearned association, in the present study we used a modified Stroop test and psychophysiological synaesthetic conditioning to further establish the genuineness of this form of synaesthesia. We compared the performance of a swimming-style colour synaesthete and a control who was trained on swimming-style colour associations. Our results showed that behavioural aspects of swimming-style colour synaesthesia can be mimicked in a trained control. Importantly, however, our results showed a psychophysiological conditioning effect for the synaesthete only. We discuss the theoretical relevance of swimming-style colour synaesthesia according to different models of synaesthesia. We conclude that swimming-style colour synaesthesia is a genuine form of synaesthesia, can be mimicked behaviourally in non-synaesthetes, and is best explained by a re-entrant feedback model.