Broiler chickens show a preference for platform perches over traditional bar perches. Further information is needed on the optimal provision of these perches in order for them to be successfully installed in commercial systems. This study was designed to compare a control and three practical levels of platform perch provision. Four commercial houses on one farm were fitted with either 1) no perches, 2) 8 perches (0.5 m2 of space per 1000 birds), 3) 10 perches (0.6 m2/1000 birds) or 4) 12 perches (0.7 m2/1000 birds). The study was repeated over four production cycles, with 28 000 mixed-sex Ross 308 broilers placed in each house. Observations of enrichment use, broiler behaviour, walking ability and litter quality were taken during weeks 2,3,4,5 and 6. Production data was also acquired post-slaughter. Perch occupancy, perching attempts and perching duration were used to assess enrichment use weekly. Broiler emotional state was determined through observations of play behaviour during a walk-through, and observations of fearfulness using avoidance testing, intensity of vocalisations (Db) and latency to approach a tray of peat. Platform perches were used fully at all levels of provision, with equal levels of occupancy, perching duration and perching attempts seen in each perch treatment (p > 0.05). Lower levels of avoidance behaviour were seen in all perch treatments compared to the control (p < 0.05), suggesting a reduced fearfulness, but there was no difference in vocalisations, latency to approach peat, or play behaviours (p < 0.05). There was no effect of treatment on average gait scores, although fewer birds were categorised with the most severe abnormalities in the highest level of perch provision. Platform perches had no effect on activity levels, production parameters or litter quality (p < 0.05). Suspended platform perches that are raised as birds age are successful at facilitating normal roosting behaviour and increase the floor space available in latter weeks. Higher levels of perch provision led to a higher level of flock roosting, and access to perches led to reduced avoidance behaviour. However, a higher level of perch provision tested did not result in an improvement to any health or welfare parameters compared to lower provision levels, and there were no clear beneficial effects of perches on activity levels or walking ability.