This paper attempts to test the recent British Industrial Relations model of trade union membership by an examination of a survey of early school-leavers in Dunedin, New Zealand which was carried out in October 1989. The findings offer strong support for the model because the same distinct strands of core motivation and remainder attitudes were evident. This demonstrates that the model could be successfully applied in a different institutional, cultural and economic context. The major cross-national differences to emerge were that most Dunedin youngsters intended to join a union; for them, collective instrumental reasons were very important and values of little significance. Furthermore, there was little evidence of disinterest or ignorance amongst the minority which was negative towards trade union membership. A recent article on trade union membership (Cregan and Johnston, 1990) suggested that conventional neoclassical theories are flawed by the free rider paradox, whereby a rational individual will not bear the costs of joining a union to gain rewards that are available to all the workforce as public goods. It proposed that the dilemma could only be solved by a membership theory which takes into account several different sources of individual motivation drawn from several disciplines. These were identified in a longitudinal survey of London early school-leavers, 1979–1981, in reasons given by young people for their membership decision, positive or negative, from which employees could be categorised in social movement parlance as core and remainder. However, the authors proposed that further direct investigations should be made in different contexts. For example, it may be that some responses were culturally or institutionally specific, or were based on economic context. Accordingly, a similar survey of a single cohort of early school-leavers was carried out ten years later in Dunedin, New Zealand. The aim of this paper, therefore, is to test the validity of the framework of the model within a different national context. The article will be organised in the following way. First, a brief summary of the Industrial Relations model of trade union membership will be presented and two hypotheses will be drawn from it. Second, the latter will be tested by a discussion of the results of the Dunedin survey and a comparison with those of the London survey. Third, implications of the findings for the consequences of the 1991 Employment Contracts Act will be briefly examined.