The Bog Ecosystem Research Initiative (BERI) project was initiated to investigate, at five climatically different sites across Europe, the effects of elevated CO2 and N deposition on the net exchange of CO2 and CH4 between bogs and the atmosphere, and to study the effects of elevated CO2 and N deposition on the plant biodiversity of bog communities. A major challenge to investigate the effects of elevated CO2 on vegetation and ecosystems is to apply elevated CO2 concentrations to growing vegetation without changing the physical conditions like climate and radiation. Most available CO2 enrichment methods disturb the natural conditions to some degree, for instance closed chambers or open top chambers. Free Air CO2 Enrichment (FACE) systems have proven to be suitable to expose plants to elevated CO2 concentrations with minimal disturbance of their natural environment. The size and spatial scale of the vegetation studied within the BERI project allowed the use of a modified version of a small FACE system called MiniFACE. This paper describes the BERI MiniFACE design as well as its temporal and spatial performance at the five BERI field locations. The temporal performance of the MiniFACE system largely met the quality criteria defined by the FACE Protocol. One minute average CO2 concentrations measured at the centre of the ring stayed within 20␘f the pre-set target for more than 95␘f the time. Increased wind speeds were found to improve the MiniFACE system's temporal performance. Spatial analyses showed no apparent CO2 gradients across a ring during a 4 day period and the mean differences between each sampling point and the centre of the ring did not exceed 10ÐObservations made during a windy day, causing a CO2 concentration gradient, and observations made during a calm day indicated that short term gradients tend to average out over longer periods of time. On a day with unidirectional strong winds, CO2 concentrations at the upwind side of the ring centre were higher than those made at the centre and at the downwind side of the ring centre, but the bell-shaped distribution was found basically the same for the centre and the four surrounding measurement points, implying that the short term (1 sec) variability of CO2 concentrations across the MiniFACE ring is almost the same at any point in the ring. Based on gas dispersion simulations and measured CO2 concentration profiles, the possible interference between CO2-enriched and control rings was found to be negligible beyond a centre-to-centre ring distance of 6 m.