Purpose – The purpose of this paper is to contribute to the debate concerning the compatibility of variable pay schemes (VPS) and trade unions, by investigating the effects of the managerial function (to increase management control over employees) and market function (to attract and retain suitably qualified employees) of VPS for trade unions. Design/methodology/approach – A critical case study approach was used to verify how the two varying functions of VPS affect the trade union of a large multi-establishment Irish financial institution. Data are based on a total of 60 in-depth interviews with key informants (trade unions, managers and non-managerial employees) at various levels in the organisational hierarchy and a representative survey of employees. Findings – While the findings confirmed the a priori expectation that the managerial function of VPS is more likely to damage unions than the market function, this study revealed that the interplay between the two functions varied over time and across outlets, and was primarily contingent on the disposition of line management towards the use of VPS and the market context. Research limitations/implications – The consequences of the shift to a managerial function of VPS following the 2008 economic crisis could only be assessed in a preliminary manner due to the time frame of the study. Practical implications – Trade unions, managers and employees need to consider the potential gaps between the intended purposes of VPS and their actual implementation to assess whether they are of potential benefit. Originality/value – This paper contributes to the existing literature in two main ways. First, it adds that the market function of VPS serves managers’ aims not only in aiding them attract, retain and motivate suitably qualified employees as indicated by Flanders (1970), but also in helping them gain great control over labour costs, which in turn, has mixed effects for trade unions. Second, in addition to the pragmatic reasons identified by previous studies, the findings show that a key determinant in unions’ acceptance of VPS involves earlier experiences of unions fighting unsuccessfully against the adoption of VPS.