The purpose of this study is to discuss the relationships between visitors' characteristics and their choice of activities, with regard to scenery and how it was perceived, in a suburban recreational forest in Japan. A visitor employed photography method was used to record the visitors' scenic perceptions. The survey was taken at Minoo Quasi National Park, a suburban area of the megalopolis city of Osaka. Eighty-one visitor groups were selected as participants of this survey. They were instructed to take photographs at their own will, of what they viewed as positive scenes, such as pleasant scenery or enjoyable events. A questionnaire about visitors' attributes and activities was also conducted to complement the visitor employed photography method. The total number of photographs analyzed was 1377. They were categorized into three types "object", "event" and "place". The participant groups were also categorized by their attributes and activities into four types. To identify the relationship between the ratio of categories of the photographs taken, and the activities of the groups, group characteristics were assessed using an analysis of variance. Furthermore, this multi-variate structure was ordinated and illustrated by canonical correspondence analysis. The results clearly indicate the differences in scenic perception and the types of activities engaged in, by size and generation of the groups and their previous experiences or familiarity with the area. In regard to the relationship between the photograph categories with group types, families with children and a groups of younger-aged parties of 3 or more, tended to take more "event" type pictures than "object" or "place" pictures. In contrast, older-aged groups took relatively few "event" pictures and concentrated mainly on "place" type pictures. Younger couples were relatively neutral, however the rate of "object" pictures taken was higher than that of any other group. The result of the canonical correspondence analysis suggests that there were sets of behavioral activities concerning the environmental and group's property. It has been thought that there are differences between visitors in their attitudes and perceptions toward the surrounding environment. These should be taken into consideration when landscape planning is established for recreational forests. (c) 2005 Elsevier B.V. All rights reserved.