This study aimed to identify hungry broiler breeders (n = 12) preferences for quantitative (control) or qualitative dietary restriction (QDR) in a closed economy environment. The QDR option was either 3 g calcium propionate/kg total feed (n = 6) or 300 g oat hulls/kg total feed (n = 6). Quantitatively restricted or QDR portions ensured equal growth regardless of choice. Birds were separately taught a Control diet versus no food and a QDR diet versus no food task to allow each diet's satiating properties to be learnt. Birds had to associate the T-maze coloured arms with dietary outcomes to immediately obtain food. Birds learnt this task easily (p < 0.001). A choice between the control diet and the QDR diet was then offered but neither group demonstrated a diet preference. Study modifications demonstrated this was not a failure to discriminate between the diets per se (the Control diet was strongly preferred under ad libitum conditions (p < 0.001)) or novel colour combination confusion (the colour associated with food was immediately selected when two novel food versus no food colour combinations were offered (p < 0.001)). Most birds still failed to show a significant preference when the Control diet quantity was increased by 50% to make it ‘obviously’ bigger and better. Therefore, it was concluded that the failure to show a dietary preference was due to task learning failure and not necessarily lack of dietary preference. Where a preference was observed it was always for the control diet. Possible reasons for this failure to learn are discussed