To examine variability in 'uptime' (the amount of time spent in the upright position).An observational study.The community (i.e., homes and schools in the Melbourne metropolitan area).Normal children between 8 and 15 years of age (18 girls and 8 boys).An 'uptimer', a lightweight, battery-operated remote monitor, was used to record uptime.Each child in the study wore the uptimer continuously over a 24-hour period, which included a typical day at school, on four separate occasions, one week apart.The variability in repeated measures of uptime in the same child (standard deviation 0.8 hours in 24 hours) was lower than the variability between the children (standard deviation 1.1 hours in 24 hours). We estimate that a single uptime measurement has reliability given by an ICC of 0.65 (95% confidence interval 0.49-0.82), which is calculated by dividing the between-child variability by the total variability (within-child and between-child variability). There were no systematic differences in mean uptime between occasions of measurement.As an estimate of the time that a child spends in the upright position, a single 24-hour period of measurement of uptime may vary considerably from one occasion to the next. However, this within-child variability is less than the variability between children. In large-scale field studies a one-off measurement of uptime may be acceptable, but in small rehabilitation studies uptime should be measured on at least three occasions.