Direct dating of the pottery is an important goal in archaeological research and many attempts have been made using radiocarbon. One important goal has been to date the earliest pottery in a region to assess the origin and dispersal of ceramic technology. Also with the increasing application of organic residue analysis to study pottery use, it has become important to combine residue data with direct dates on the artefact being investigated. In this study we have radiocarbon dated different organic materials associated with archaeological potsherds from three Ertebølle sites in Northern Germany. These are hunter/fisher/gatherer sites located in coastal or riverine setting and correspond to some of the earliest pottery bearing sites in this region. We have chosen to radiocarbon date different fractions on the pottery including “foodcrusts” (charred deposits from the inner surface of sherds), “soot” (charred deposits from the outer side of sherds), plant remains from inside the clay matrix, and lipids extracted from the ceramic matrix. All of these are potentially problematic media for AMS dating: ‘Foodcrusts’ and absorbed lipids can appear too old because of the marine or freshwater reservoir effect, such as when aquatic products have been prepared in the pottery. Soot can derive from old wood that was used for the hearth fire, or from (potentially aquatic) food that boiled over. Plant remains may have been present in the clay for a long time before manufacture of the pottery. Post-depositional contamination with organic carbon, such as humic acids, may also be problematic. We present these data with radiocarbon datings of contemporaneous terrestrial and aquatic samples to find out the true age of the pottery and estimate the reservoir age. Lipid analysis and bulk carbon and nitrogen stable isotope measurements are used to give additional information about the former contents of the pottery and the risk of reservoir effects.