The ranging behaviour of broiler chickens kept in free-range housing systems remains poorly understood, despite access to the outdoor range being their main feature. We investigated the impact of allowing chickens to have range access on both sides vs. one side of the shed, using 24 flocks of approximately 40,000 Ross 308 chickens of mixed sex on one commercial farm across winter and summer. Sheds were identical and pseudo-randomly allocated to either double-sided (no modification) or single-sided (by keeping one side closed at all time) range access treatment. Flocks were first provided with range access from 15 to 17 and 21 to 27days of age for summer and winter flocks, respectively. Live outdoor observations were conducted daily for the first week after first range access and every other day from the second week onwards until the day prior to depopulation (44days of age), twice daily in each morning and evening during anticipated peaks of range use. Indoor side fidelity was also assessed by spray marking 320 chickens in total on the right- and left-hand side of the shed, one colour each side, and conducting two to four repeated counts of the colour-marked individuals in each location over the week following marking. Indoor count results showed that colour-marked chickens had 50% chance of being found on either side of the shed (P<0.001 from a side preference), and therefore did not support the hypothesis that chickens show indoor side fidelity in commercial conditions. Consequently, we could not elucidate whether an individual chicken would cross to the opposite side of the shed to access the range. Winter flocks had infrequent range access and low number of birds on the range (49±175 chickens outdoor at any one time). For summer flocks, the ANOVA model explained 72.5% of the variance, with the number of chickens observed on the range being affected by the interaction of treatment and age (P<0.001); more chickens were observed on the range when range access was available on both sides of the shed compared to a single side, from the seventh day of range access onwards. Hence, shed design can limit ranging in broiler chickens during the period of high ranging activity. For double-sided sheds, the number of chickens on the range steadily increased until an average of 28% of the flock could be seen on the range at one time, by 44days of age prior to depopulation.