A large part of the world’s inland waters, including streams, rivers, ponds, lakes, and reservoirs is subject to occasional, seasonal, or even permanent drying. Moreover, the intermittent inland waters are increasing in many areas of the world because of climate change, anthropogenic influence, and land-use change. However, information on the carbon dioxide (CO2) fluxes from intermittent inland waters is scarce. Thus, a comprehensive assessment of CO2 emissions in intermittent inland waters is necessary and is a research gap in China. Here, we review the current knowledge on CO2 fluxes from lotic (streams and rivers) and lentic (ponds, lakes, and reservoirs) inland waters during dry and rewetting phases, considering controls and sources as well as implications of the two phases for local and global scale estimates. In exposed sediments of dry riverbeds, it can be a long time between wet and dry conditions. At the onset of drying, the increased oxygen availability can stimulate enzymatic activities and overall microbial growth, leading to an increased breakdown of organic matter and the subsequent release of CO2. As in any ecosystem, carbon emissions derive from a set of abiotic and biotic sources, the relative importance of which is tied to intrinsic features, such as geology, microbial communities and available organic substrates, and environmental controls, such as moisture and nutrient availability. Besides microbial respiration, geologic carbon can be a relevant source of CO2 emissions. Particularly in carbonate-rich regions, photodegradation of organic matter on dry light-exposed sediments can also lead to increased CO2 losses. In addition, covered vegetation metabolism, including photosynthesis and respiration, can be a relevant part of dry riverbed sediment CO2 emissions. Short rewetting episodes can trigger microbial respiration and CO2 release from the exposed sediments, and marked increases in the CO2 flux from dry sediments are to be expected during rewetting after a dry period, similar to soils, and it is known as the Birch effect. Immediately after rewetting, gases in pore spaces are displaced by water, potentially causing a significant efflux of CO2, impacting the overall annual CO2 flux. Rewetting events can also cause the remobilization of organic matter and nutrients. At the micro-scale, pore-filling will increase the dissemination of water-soluble substrates and mobilize labile substances from dead biomass. Aggregate disruption and the destabilization of organic-mineral complexes will also liberate carbon otherwise protected from microbial action. Other major drivers of CO2 fluxes from dry riverbeds include the presence or absence of vegetation or micro-phyto benthos in the exposed sediments, sediment texture, sediment temperature, and sediment organic matter. The organic matter quantity may be more important than the quality in controlling the magnitude of CO2 fluxes from the exposed sediments. Our conservative estimates indicate that adding emissions from intermittent inland waters to current global estimates of CO2 emissions from inland waters could result in an increase of 0.51 Pg C a−1 or ~1/4 of total fluxes. Therefore, intermittent internal waters should be considered when estimating the CO2 emissions of global inland water. Future investigations should focus on the in-depth study of the main sources of uncertainty involved in carbon emissions from intermittent internal waters to understand the physical and biological mechanisms behind this flux, and the impact on the global carbon cycle. Important study topics include: (1) Accurate estimation of the area of intermittent inland waters, in particular, the area of small-flow rivers, scattered ponds, and the area of headwaters; (2) CO2 efflux measurement of dry riverbeds, including the determination of CO2 efflux directly and indirectly, and combining the two methods; (3) the mechanism of CO2 release from dry riverbeds and factors driving CO2 fluxes from intermittent inland waters, such as the ratio of autochthonous and allochthonous carbon in sediments during dry and rewetting episodes; and (4) deep understanding of the changes in extension and seasonality of intermittent internal waters, particularly those in response to climate oscillations, land-use changes, and potential feedback related to climate at decadal time-scales. The results from these additional studies will help explain completely the inland water carbon emissions and identify potential implications for regional and global carbon cycles.