The northern permafrost regions have been estimated to contain approximately 1700 Pg of organic carbon, about 90% of which occurs in permafrost deposits, representing approximately 50% of the estimated global below-ground organic carbon stock [Tarnocai et al., 2009]. In a warming climate part of this large carbon stock could be released as carbon dioxide (CO2) or methane (CH4), thus generating a positive feedback to climate change [McGuire et al., 2009]. Inland waters have been recognized as important participants in the carbon cycle, actively processing the carbon derived from terrestrial ecosystems that then makes its way into the atmosphere, oceans, and sediments [Tranvik et al., 2009]. Most of the research into this aspect of the carbon cycle has addressed water bodies with a surface area of several hectares [Kling et al., 1992; Hamilton et al., 1994; Duchemin et al., 1999; Jonsson et al., 2003; Åberg et al., 2004; Repo et al., 2007; McGuire et al., 2009], while ponds with surface areas of only a few square meters and depths measured in decimeters have received scant attention [Boike et al., 2008; Laurion et al., 2010] and are not taken into account for global estimates of CO2 emissions because they are invisible to most satellites [Muster et al., 2012]. However, substantial carbon emissions have been observed from small and medium-sized lakes in both sub-Arctic [Hamilton et al., 1994; Jonsson et al., 2003; Huttunen et al., 2002a, 2003; Repo et al., 2007] and Arctic environments [Kling et al., 1992; Blodau et al., 2008; McGuire et al., 2009; Shirokova et al., 2009]; a summary of CO2 emissions from surface waters can be found in Table 1. Surface waters in Arctic and sub-Arctic environments emit on average 0.5 g CO2-C m−2 d−1 (ranging between 0.0 and 3.0 g CO2-C m−2 d−1, with a standard deviation of 0.7 g CO2-C m−2 d−1) [Kling et al., 1992; Hamilton et al., 1994; Duchemin et al., 1999; Cole et al., 2000; St. Louis et al., 2000; Huttunen et al., 2002a, 2003; Åberg et al., 2004; Repo et al., 2007; Blodau et al., 2008; Shirokova et al., 2009] (Table1). Extrapolation of these fluxes to larger areas or over longer time periods remains challenging, however, because the total surface area of ponds that are invisible to satellites is virtually unknown and emissions during spring thaw and autumn freezeback periods are poorly constrained. This study aimed to fill the gap in our knowledge concerning the role of small ponds in the Arctic carbon cycle. The overall objectives were (i) to quantify the organic and inorganic carbon fluxes into/from water bodies of various sizes, and (ii) to evaluate their contributions to the larger scale CO2 budget of a tundra region encompassing several square kilometers. The research was carried out in a typical tundra landscape on Samoylov Island, in the Lena Delta of northeastern Siberia, during the late summer and early freezeback of 2008.