An analysis of five different cohorts spanning a thirty-year period reveals little support for any particular model of church attendance as related to chronological aging. Although cohorts share trends in church attendance by sex and specific dates or time periods, each cohort seems to manifest its own distinct pattern as related to age itself. These findings suggest that even though church attendance varies by age, other factors such as specific cohort membership and general societal pressures may be more meaningful explanatory factors. A recurrent topic in the fields of both gerontology and sociology of religion is the relationship between church attendance and age. However, in spite of abundant research conducted on this topic, many questions remain largely unresolved, and a number of contradictory schools of thought still exist. Howard M. Bahr (1970) presents comprehensive descriptions of four models which have been used to classify research findings on the relationship between aging and church at-, tendance.1 He calls these four models traditional, stability, family-cycle, and disengagement. The traditional model views church attendance as reaching a low between the ages of 30 and 35, then increasing steadily until old age. The stability model sees church attendance as unrelated to age and as being fairly constant throughout one's lifetime. The family-cycle model suggests a relationship between age and church attendance depending on age-linked marriage and child rearing stages. According to this model, the peak in church attendance appears among young parents assumed to be concerned about their children's religious education. Finally, the disengagement model implies a gradual decline in church attendance from middle to old age. As the person ages, he gradually reduces his church-related activity along with other behavior patterns considered inappropriate for the aged in our society. The research of Cauter and Downham (1954), Fichter (1952; 1954), Glock et al. (1967), O'Reilly (1957), Smith (1966), as well as the writing of Argyle (1959) support the traditional model. That of Catholic Digest (1953; 1966), Lazerwitz (1961), Orbach (1961), and to some extent Streib (1965) upholds the stability model. Albrecht (1958) seems to be the chief exponent of the family life-cycle model, although Lazerwitz (1961) did find increased regularity of church attendance among Protestants with children five years old and older. The disengagement model finds support from the works of Barron (1958; 1961), Catholic Charities of St. Louis (1955), Hunter and Maurice (1953), McCann (1955), and Mayo (1951). A major source of confusion seems to arise from the two major methodological strategies used by the researchers noted above. We suspect that the existence of four empirically tested, yet contradictory models is due to the use of methodologies inappropriate for the problem. With the exception of a few restricted longitudinal studies (Streib, 1965), most of the research in this area can be classified as either cross-sectional or retrospective in nature. Crosssectional analyses assume that the activity pattern of those in younger age categories accurately portrays that of older people when they were younger. Such an approach ignores changes in behavior due to intergenerational differences as well as the possible existence of social change taking place during the lifetimes of respondents. In short, the cross-sectional approach presents a static view of behavior which often results in erroneous conclusions (Geriatric Focus, 1973). The retrospective approach attempts to approximate life-cycle patterns in church at* The authors thank the Roper Public Opinion Research Center, the University of Georgia Library System for making available the data used in this paper, and the University of Richmond which financed typing of the revised manuscript. 1 For a more complete summary and discussion of studies in this area, see Moberg (1965) and Orbach (1961).