INTRODUCTIONIt has been proposed that leisure behavior is a dialectical process, according to which an individual seeks both stability and change, structure and variety, familiarity and novelty in his or her intrapersonal and interpersonal encounters throughout life cycle (Iso-Ahola, 1980a). In a similar vein, other leisure researchers (Howe, 1988; Kleiber & Kelly, 1980; Kelly, 1987; Mobily, 1987; Osgood & Howe, 1984) have emphasized life-span approach and notion of continuity and change in understanding leisure behavior. Hence, human development is seen as a process of continuity and change, suggesting that the leisure self' develops throughout life, adapting and renewing itself (Mobily, 1987). The tendency toward both novelty and stability (and thus optimal arousal) in leisure behavior (Iso-Ahola, 1980b, 1989) raises a question about how to balance need for continuity and change, variability and routine in leisure. One way to do this is to start new activities or replace old activities with new ones while continuing to participate in some old and familiar activities. several studies have recently explored this issue (Iso-Ahola & Hayllar, 1991; Jackson, 1990; Jackson & Dunn, 1988; McGuire, O'Leary, Yeh, & Dottavio, 1989). The present study extends these efforts and seeks to determine patterns of starting, ceasing, and replacing leisure activities over life-span.Patterns of leisure behavior can be studied at both macro and micro levels. The former focus on determination of general patterns of starting, ceasing and replacing leisure activities at aggregate and sub-group levels of population. The latter, on other hand, refers to intrapersonal patterns of change in leisure behavior. That is, people can be classified into replacers, adders, quitters, and continuers Jackson & Dunn, 1988; McGuire et al., 1989) in relation to actual change or non-change that individuals have made in their leisure behavior over a given time period--in this instance, one year.(1) These two sources of evidence are used in present study to examine patterns of leisure behavior across life-span.BACKGROUND TO THE STUDYLEISURE LIFESTYLESThe study of patterns of leisure behavior (starting, ceasing and replacing) over life-span is important from both individual and societal standpoints. For one thing, such research helps shed light on versus passive leisure lifestyles and their impact on person and society. Economists (Keeler, Manning, Newhouse, Sloss & Wassermann, 1989) have shown that lifetime external costs of relatively inactive persons ($1,900) are greater than those of smokers ($1,000) but less than those of heavy drinkers ($4,600). Such costs suggest that prevention of a sedentary lifestyle is an important individual and societal issue (Kirshenbaum, 1987) and provides an economic rationale for health-promotion programs and leisure lifestyles.Besides its economic benefits, an leisure lifestyle seems to have positive effects on human health (Chalip, Thomas & Voyle, 1992; Coleman & Iso-Ahola, 1993; Iso-Ahola, 1994; Taylor, Sallis & Needle, 1985). For example, Casady (1975, p. 138) has noted that during mainstream years from 20 to 50, those who are happiest and best adjusted are participants in life. They are intellectually alert, socially assertive, engaged with environment and other people. Similarly, Flanagan (1978) found in his nation-wide survey that active recreation was one of six areas showing largest correlation coefficients with indicators of a high quality of life. In addition to mental health benefits (Hull, 1990; Morgan & O'Connor, 1988; Stephens, 1988; Wankel & Berger, 1990), leisure lifestyle has been shown to be beneficial also from physiological standpoint (Blair, 1988; Folsom et al., 1985; Ulrich, Dimberg & Driver, 1990).Research in this area has led Diamond (1984) to her famous axiom, use it or lose it. …