Background: ActiGraph is widely used to measure physical activity in field studies. However, there is limited research on the validity in older adults. As the metabolic cost of ambulation is greater in older adults and their gait can be different, this research is needed. Therefore, the aims of this study are (1) to identify the best placement for the accelerometer (ACC) and (2) to create novel equations to predict intensity of physical activity with ActiGraph. Methods: 20 participants (9 female, 11 male) performed 20 minutes of treadmill walking with increasing speed of 0.5 km/h every 4 minutes (3.0 to 5.0 km/h) and five typical daily activities (reading newspaper, shopping, brushing, cycling and aerobics) in older adults for 4 minutes each. Indirect calorimetry was used to validate the ActiGraph estimates of physical activity intensity. ACC were worn on both ankles, both wrists and at the hip. Included were cardiorespiratory healthy Scots aged 55 – 75 years living near Glasgow. Bland-Altman-Plots, Pearson's correlation and linear regression were used to analyse the data. Results: Participants were on average 62.9 years old (SD 3.6) with a mean BMI of 26.7 kg/m2 (SD 5.4). With the treadmill walking, ACC measurements performed reasonably well, irrespective of placement. However, with daily activities the performance is dependent on activity type and placement, e.g. cycling was best measured on the ankle and brushing on the wrist. Measuring accuracy was dependent on age (p < 0.01) and sex (p < 0.001) for ankle placement. Age and sex did not affect the measuring accuracy of hip or wrist placement. In daily and treadmill activities is a high positive correlation between dominant and non-dominant ankle (daily: r = 1.00; treadmill: r = 0.96) and both wrists (daily: r = 0.95; treadmill r = 0.85). Conclusion: In daily activities, the best placement of the ACC depends on the type of activity. It is irrelevant, whether the device is worn on the dominant or non-dominant wrist or ankle.