Two related experiments involving broiler chickens are presented. Experiment 1 evaluated the use of six perch types: (1) an A-frame design (incorporating a platform and ramps), (2) a ‘flat top’ ramp, (3) a curved ramp, (4) a suspended bar, (5) a fixed bar and (6) a suspended platform. Two of each perch type was provided in one commercial house on each of two farms over two 6-week production cycles. Each perch was videoed for a 24 h period in weeks 1–6 of cycle 1, and weeks 1–5 of cycle 2. Scan sampling was used to assign an occupancy score to different perch components (platform, bar and ramp, where appropriate), and an overall weighted occupancy score also calculated. Counts were made of perching and failed perching attempts following selected scans in cycle 1. There were significantly higher occupancy scores for platform than for bar or ramp components, and this was apparent across the production cycle. This resulted in a higher overall weighted occupancy score for suspended platforms. The percentage of failed perching attempts was significantly greater with fixed and suspended bar perches than with the curved ramp. Three treatments were assessed in Experiment 2: (1) provision of six suspended platform perches (P), (2) provision of six suspended platform perches and four peat-filled dust baths (PD), and (3) control treatment with no platform perches or dust baths. Treatments were applied in one of three houses on each of two farms, and replicated over three cycles. Two perches in each of the P and PD treatments were videoed for a 25 min period in weeks 3, 4 and 5, and number of birds using the perches recorded. The severity of angular leg deformities, hock burn and pododermatitis lesions, and walking ability were scored in weeks 3, 4 and 5, and prevalence of pododermatitis and hock burn recorded at slaughter. Litter moisture and production-related measures were also taken. On average, 26 birds (12.6 birds/m2) occupied the perches, and this was not affected by provision of dust baths or age. Treatment did not significantly affect any of the measures taken. It is concluded that broilers prefer to perch on platforms rather than bars or ramps, and thus that platforms better cater for an important behavioural need. However, provision of platform perches, even in combination with dust baths, did not improve leg health, and future research should investigate greater levels of provision of these enrichments.