Purpose The purpose of this paper is to compare the impact of different multi-unit promotions (MUPs) and a single-unit promotion (SUP) on store-level sales and consumer-level purchase probability and quantity decision. Design/methodology/approach The paper combines two empirical studies. Study 1 applies a hierarchical multiplicative model to store-level sales data for four product categories provided by a large Dutch retail chain. Study 2 presents a laboratory experiment in which the quantity requirements of the two focal MUP frames are manipulated to assess their impact on consumer purchase decisions. Findings The paper provides empirical evidence for the superiority of the “X for $Y” above “X + N free”, which confirms the hypotheses based on prospect theory, mental accounting and theory about gift-giving. Quantity requirements of four to five units show the largest effects. In addition, the superiority of the “X for $Y” frame holds for functional product categories, but not for the hedonic categories. Practical implications The paper provides managerial insights into the relative effectiveness of alternative MUPs and an SUP and how this promotional effectiveness depends on the type of product category and quantity requirements. Originality/value This paper combines actual sales data and experimental data. This “mixed approach” extends existing knowledge by comprehensively evaluating two MUP frames, namely, “X + N free” and “X for $Y” promotions, and an SUP.