Practical strategies to alleviate the negative effects of feed restriction on fast growing standard broiler breeders (S) were studied. A dwarf heavy broiler breeder selected for better viability and reproductive traits at the partial expense of growth experimentally (E) was compared to S. A slow growing dwarf broiler breeder used for Label chicken production (L) was used as a control. There were four treatments: S fed ad libitum (SA), S feed restricted (SR), E and L fed ad libitum from 0 to 40 weeks of age. Chicks were distributed in 48 pens of 14 birds with two marked focal birds per pen. Time-budgets of the 96 focal birds were measured over 3 weeks starting at 3, 9, 15, 21, 27 and 33 weeks of age, by scan sampling (12 scans/h) during 3 periods of I h (prior to, immediately after and 5 h after feed distribution). During the growing stage (3-17 weeks of age), in Period 1, E spent more time resting, SA preening and SR standing and pecking at the feeder than L. In Period 2, the major activity of SR was to eat and drink, whereas ad libitum-fed birds rested more often. In Period 3, SR substituted the time spent eating by birds of the other treatments, by pecking at the empty feeder (25%) or at the litter (20%). Average standing plus resting time during the growing stage were 44.6% for L, 37.6% for SA, 34% for SR and 50.1% for E. The E genotype had a better egg production and liveability than SA and consistently rested more. During the laying stage, results were similar except that sexual maturity was delayed by 6 weeks for SR and less severe feed restriction induced less total pecking behaviours. Taken together, these comparisons suggest that the behavioural characteristics of each genotype of broiler breeder is more important than the effect of feed allowance and/or selection for growth rate per se. In contrast to comparisons between treatments, the correlations between individual birds or pen means over time were low. (c) 2005 Elsevier B.V. All rights reserved.