Objective To explore the characteristics associated with active travel, including leisure activity, and to assess the contribution of active travel to and from school to children's overall activity level.

Design Cross-sectional health examination survey.

Setting The Health Survey for England 2008 surveyed a random nationally-representative sample of the general free-living population in England.

Participants Children aged 5–15 years who had complete information on all variables used (n=4468), including self-reported physical activity. Analyses predicting objectively measured physical activity were repeated using a subsample of children with at least 4 days actigraph wear (n=603).

Main outcome measures Active travel (walked to/from school at least once in the last week, or cycled to/from school at least once in the last week, vs neither); met the physical activity recommendations (at least 60 min daily on all 7 days vs did not, self-reported); actigraph-measured activity tertile (highest, middle, vs lowest); other walking or cycling (did any other walking, did any other cycling).

Results 64% of children walked to/from school, 3% cycled, and 33% did neither; 410 children with sufficient actigraph wear walked and 19 cycled to/from school at least weekly. Children who walked or cycled to/from school were more active than those who did neither (self-reported, met recommendations: walking OR 1.17, (95% CI 1.00 to 1.37), cycling OR 1.93 (1.33 to 2.80); actigraph-measured highest activity tertile: walking OR 3.51, (1.81 to 6.80), cycling OR 5.22, (0.90 to 30.38); actigraph-measured middle activity tertile: walking OR 2.09 (1.21 to 3.6), cycling OR 6.54 (1.54 to 27.80)). The profile of walkers and cyclists differed. Walkers generally came from a deprived area (OR 1.45, 1.25 to 1.69) and were less likely to have a limiting illness (OR 0.69, 0.54 to 0.87). Cyclists were generally older (OR 1.12, 1.06 to 1.19), male (OR 4.03, 2.60 to 6.25), and most likely to meet the recommendations (self-report OR 1.86, 1.28 to 2.70). For self-reported activity, time spent cycling to/from school (OR 1.31, 1.09 to 1.59) contributed more to meeting the recommendations than time spent walking to/from school (OR 1.08, 1.02 to 1.15) or in sports (OR 1.17, 95% CI 1.14 to 1.20). For actigraph-measured activity, only time spent walking to school (OR 1.80, 1.41 to 2.30) and in sports (OR 1.10, 1.01 to 1.20) were significantly associated with being in the highest activity tertile. However, the actigraph measures bi-axial movement so does not capture cycling.

Conclusion Children who reported walking or cycling to/from school were more active than those who did neither. Longitudinal studies are required to ascertain whether encouraging active travel increases overall activity levels in less active children.