Recent evidence suggests that between-individual variation in posture and posture transitions may have important health consequences in adults. The early life development of between-individual variation in posture and posture transitions has not been studied, and the physiological consequences of such variations in childhood are unknown, largely because of the absence of objective methods for measuring posture and posture transitions in young children. This study aimed to examine the objective measurement of posture transitions in pre-school children with the activPAL™ monitor (PAL Technologies, Glasgow). Single-unit activity monitors such as the activPAL™ have a limited output, with data categorized as 'sit/lie', 'stand' or 'walk' and the consequences of this for measurement of posture transitions in young children are unknown. Thirty children (mean age 4.1 years) were videoed for 1 h in nursery while wearing an activPAL™. Video was analysed on a second-by-second basis, with all postures categorized. From direct observation, time spent was sit/lie 46%; stand 35%; walk/run 16%; 3% was spent in heterogeneous non-sit/lie/upright postures (crawl, crouch, and kneel up). Despite these 'non-standard' postures being responsible for a low proportion of time, posture transitions involving them contributed to 34% of total transitions. There was a significant rank-order correlation (r = 0.79, p < 0.0001) between the number of posture transitions measured by activPAL™ and by direct observation. 'Non-standard' postures in young children are probably not a problem if the aim is to measure total time sedentary or active, and the activPAL™ may measure between-individual variation in transitions adequately in young children. However, non-standard postures may present problems for the detailed characterization of posture transitions in early childhood.