The church-sect typology, originating in the work of Troeltsch and Max Weber, has been used, criticized and re-formulated with varying success by many sociologists interested in religious phenomena. Actually there is no generally agreed upon typology of religious bodies which can be distinguished as the church-sect typology. Rather there are numerous variations on a theme; a research tradition which places 'church’ at one end of a continuum and “sect” at the other. This study aims first at drawing together ideas from the wider methodological literature on types and typologies and from the church-sect tradition, allowing each to play upon and hopefully to illuminate the others. This interplay of methodological ideas occurs in chapters I and II. A second aim is to test rapidly but consistently a number of church-sect hypotheses arising from the literature against quantitative data obtained by questionnaire from members of religious groups. In chapter III the content of the questionnaire is discussed and an outline of the background characteristics of the respondents is given while in chapter IV various hypotheses are set up and tested. Chapter IV concludes with the question, what other ways are there of describing and classifying religious groups? The chief alternative to the typology is description in terms of dimensions. In chapter V the statistical technique of factor analysis is used in conjunction with available data in an attempt to discover one or more empirically justifiable and theoretically meaningful dimensions upon which individual groups may be located and over which any number of religious groups may be compared. Throughout, this study is primarily exploratory; it attempts to examine potentially useful approaches and methods rather than to argue the case for a specific scheme. By and large substantive findings are not emphasized because they are not generalizable. The data on which they are based are not sample data and statistical inference does not enter the study except peripherally in chapter IV. A broadly based definitive work using some of the approaches suggested here must wait until the writer or another researcher is able to collect sufficient data sampled adequately from a wider population.