Brown Shaver hens responded on a concurrent chain procedure, containing two independent initial link keys, that ran on VI 30-s schedules. Following the conclusion of the initial link schedules either 1 or 2 terminal link keys were made available to the subject, all of which ran on a FI 20-s schedule and resulted in 3-s access to reinforcement. The results replicated Catania’s (1975) findings demonstrating preference for the terminal link which contained two available keys. The second experiment aimed to equalise terminal link entry ratios by changing the initial links to concurrently running dependent VI 30-s schedules. The last experiment attempted to increase the observed preference for the terminal link that contained two keys, by reducing the initial link schedules to dependent VI 10-s. These later experiments were not able to successfully replicate the strength of the preference for the 2 key terminal link shown in the initial experiment and by Catania (1975).