PurposeThe purpose of this paper is to report the results of two experiments designed to examine the effect on consumers of the way in which price discount messages are expressed, or “framed”.Design/methodology/approachBoth studies involved stated‐preference choice modelling experiments. The aim of the first experiment was to test the hypothesis that a price reduction framed in dollar terms is more effective for high‐priced items, whereas a price reduction framed as a percent discount is more effective for lower‐priced items. The aim of the second experiment was to determine which of four alternative ways of expressing the same 33 per cent price discount – cents off, percent discount, or one of two volume discounts – is most effective.FindingsFor two “low‐priced” items, potato chips and cola drinks, the framing of a price discount had little or no effect. However, for two ”high‐priced” items, stereos and computers, framing a discount in dollar terms was significantly more effective than expressing it as a percent off discount. For three fast moving consumer goods the most effective framing of the same price discount depended on whether the product concerned was amenable to stockpiling. For tinned spaghetti, which is relatively cheap and easy to store, volume discounting was more attractive than a monetary discount, whereas for bottled water and semi‐soft butter, which are more expensive and bulkier, the opposite was true.Originality/valueFor high‐priced products, it is better to express price discounts as dollars or cents off than as a percentage off; the opposite may be true for low‐priced products, but this is much less certain. However, if using a volume promotion, “buy x get one free” is likely to be more effective than “y for the price of x”.