Elucidating the effect of source-sink relations on berry composition is of interest for wine grape production as it represents a mechanistic link between yield, photosynthetic capacity and wine quality. However, the specific effects of carbohydrate supply on berry composition are difficult to study in isolation as leaf area or crop adjustments can also change fruit exposure, or lead to compensatory growth or photosynthetic responses. A new experimental system was therefore devised to slow berry sugar accumulation without changing canopy structure or yield. This consisted of six transparent 1.2 m chambers to enclose large pot-grown grapevines, and large soda-lime filled scrubbers that reduced carbon dioxide (CO) concentration of day-time supply air by approximately 200 ppm below ambient.In the first full scale test of the system, the chambers were installed on mature Shiraz grapevines for 14 days from the onset of berry sugar accumulation. Three chambers were run at sub-ambient CO for 10 days before returning to ambient. Canopy gas exchange, and juice hexose concentrations were determined. Net CO exchange was reduced from 65.2 to 30 g vine day, or 54%, by the sub-ambient treatment. At the end of the 10 day period, total sugar concentration was reduced from 95 to 77 g L from an average starting value of 23 g L, representing a 25% reduction. Scaling to a per vine basis, it was estimated that 223 g of berry sugars accumulated under ambient supply compared to 166 g under sub-ambient, an amount equivalent to 50 and 72% of total C assimilated.Through supply of sub-ambient CO using whole canopy gas exchange chambers system, an effective method was developed for reducing photosynthesis and slowing the rate of berry sugar accumulation without modifying yield or leaf area. While in this case developed for further investigations of grape and wine composition, the system has broader applications for the manipulation and of study of grapevine source-sink relations.