Participant compliance is an important issue in studies using accelerometers. Some participants wear the accelerometer for the duration specified by the researchers but many do not. We investigated a range of demographic factors associated with participant compliance in obtaining analyzable accelerometer data. A total of 3601 participants (aged 47.6±13.1 years, 44.6% male) were included. They were asked to wear an accelerometer (ActiGraph) for four consecutive days after completing a household survey during March 2009-January 2011 in Hong Kong. Participants wore the accelerometer on average for 13.9h in a 24-h day. No significant difference was found between males and females (p=0.38). Using log-linear regression, it was found that older participants (0.5% more wearing hours for each year of age, p<0.001), those with full-time job (p<0.01), with tertiary education (p<0.01), non-smokers (p<0.01) and with high self-reported health (p<0.05) wore the accelerometer for more hours. These results provide details for estimating compliance rates for samples with different characteristics and thus sample size calculation to account for participant compliance.