The possibility of using mobile communication devices for travel activity survey is discussed. The PHS based system gives the precise location of an individual. Through the field test in Osaka, it is found that a person will be at the calculated position within 180 meters of the accuracy with 90% of confidence. A rule-based procedure of transforming the positioning data to the activity diary of individuals is developed. A series of individual data is calculated in the conditions of the maximum moving distance and the minimum stopping time. The estimated time-space path in a day is compared with the travel diary obtained by the questionnaire survey of the conventional form. It is found that the estimated travel diary using the PHS data is consistent with the diary of the conventional survey form. The mobile communication systems would be available in a supplemental survey of a large scale travel behaviour survey.