Our objective was to investigate whether grape berry growth and colour development is associated with variation in abscisic acid (ABA) levels during ripening under an irrigation regime of partial rootzone drying (PRD). The study was carried out using container‐grown Tempranillo grapevines (established as fruit‐bearing cuttings) with a split root system (occupying two contiguous pots) and grown under controlled conditions. Three irrigation treatments were imposed – a control (well watered), and two forms of deficit irrigation, viz. partial rootzone drying (PRD), and sustained deficit irrigation (SDI). Under SDI, a given volume of irrigation water was applied uniformly and simultaneously to both sides of the split roots of each vine. Under the PRD regime, the same total volume of irrigation was applied, but as separate allocations to each side of the split root system in turn, and alternating on a 10‐day cycle. Because both deficit treatments received the same total amount of irrigation water (about 50% of the volume applied to controls), we were able to distinguish between the effects of deficit irrigation per se, versus any specific impact PRD might be having on vine performance. Outcomes were as follows. Compared to well‐watered control vines, yield per plant, and weight per bunch were both reduced significantly under SDI, but were sustained close to control values under the PRD regime. SDI and PRD thus differed significantly in their respective impacts on vine performance, and comparing just these two forms of deficit irrigation, PRD resulted in greater yield, higher bunch weight, bigger leaf area and increased berry weight compared with the SDI treatment. Moreover, PRD also induced greater accumulation of skin anthocyanins at harvest, compared to SDI. Berry ABA concentration increased continuously throughout veraison, achieving a maximum at the end of this period. The increase of berry ABA concentration was earlier and faster in PRD than in SDI, which exhibited only a slight increase in berry ABA by the end of veraison. These distinctive responses to PRD compared to SDI imply that the alternating wet‐dry cycles of PRD, rather than simply a deficit irrigation, as in SDI, had a distinctive effect on growth, ripening and berry composition. Our data imply that these different responses might be related to ABA physiology, and especially to ABA levels in berries during ripening.