For contacting a virtual cylinder located near the centre of the experimental tank, 4 male Siamese fighting fish (Betta splendens) received food reinforcement on a fixed-ratio (FR) 6 schedule. Ten times per second a tracking system recorded, in 3 dimensions, the location of the fish in the experimental tank. For each fish, the number of responses per session, the 2 dimensional swimming patterns, and the response patterns as observed in cumulative response plots were analyzed in 4 experimental conditions. The experimental conditions were: (a) both a response-location stimulus and a response-feedback stimulus were present; (b) neither a response-location nor a response-feedback stimulus were present; (c) only a response-location stimulus was present; (d) only a response-feedback stimulus was present. After experiencing the conditions, 2 fish were exposed to extinction, first with only the response-location stimulus present and then with the addition of the response-feedback stimulus. The combination of the response-location and the response-feedback stimulus and the response-location stimulus alone resulted in the highest number of responses per session and swimming patterns more concentrated in and around the target response location. The absence of a response-location stimulus and a response-feedback stimulus resulted in fewer responses per session and swimming patterns more spread out than in the conditions where the response-location stimulus was present. The response-feedback stimulus alone produced patterns similar to the condition where neither stimuli were present. In neither condition was an FR break-and-run response pattern observed in the cumulative response plots. The addition of the feedback-stimulus to the extinction resulted in a temporary increase in responding for one fish, but not for the other. The results suggested that the response-location stimulus functioned as a discriminative stimulus and that the response-feedback stimulus functioned as a conditioned reinforcer when combined with the coin, though its role in the absence of the coin was less clear. The characteristics of the response topography and the variability in the location of the consumption response may have contributed to the lack of a break-and-run pattern under FR6 with fish.