The aim of this study was to compare physical activity prevalence estimates among the International Physical Activity Questionnaire (IPAQ), steps/day, and accelerometer in a sample of government employees in Kangar, Perlis, Malaysia. Ten government agencies in Kangar were randomly chosen, and all employees were invited to participate. A self-administered questionnaire was employed to obtain information on socio-demographic characteristics and a physical activity assessment using the IPAQ. Anthropometric measurements, which include measurements of weight, height, body mass index, percent body fat, waist and hip circumference, were carried out. An accelerometer was used to assess total daily energy expenditure and the number of steps/day. A total of 272 respondents were involved in this study with a response rate of 83.2%. According to IPAQ, accelerometer and steps/day, the majority of the respondents (22.0%, 55.1%, and 77.6%, respectively) were classified as sedentary. The agreement between physical activity level as determined by the accelerometer vs. the IPAQ (Kappa=-0.46 {95% CI -0.384,-0.536}, p=0.238) and the IPAQ vs. steps/day (Kappa =0.037 {95% CI 0.090,-0.016}, p=0.175) was not significant, but the agreement between physical activity level as determined by the accelerometer vs. steps/day was classified as fair (Kappa=0.296 (95% CI 0.392, 0.200}, p<0.001). Our study highlights the need for a valid, accurate, and reliable self-report physical activity assessment tool for Malaysian adults.