Using the TANDY AMS facility (600 kV) at ETH Zurich the seawater-derived (authigenic) Be-10/Be-9 ratio of marine sediment samples is measured without the addition of Be-9 carrier. This novel method reduces systematic uncertainties because the Be-10/Be-9 ratio of a sample is determined in only one (AMS) measurement. A challenge of carrier-free AMS is to avoid any contamination of the sample with Be-9 during the chemical preparation. Further, the leaching procedure has to be reproducible and ideally should attack the authigenic Be of the sediments only, leaving the detrital Be untouched. The low amount of stable Be-9 in the unspiked samples causes low currents during the AMS measurement. This requires a good stability and sensitivity of the AMS setup. Our first results show that the new preparation method is reliable and that background from stable Be-9 is avoided. For a comparison study, sediment samples from two cores located in the Arctic Ocean (HLY0503-09JPC, HLY0503-14JPC) were used. The authigenic Be-10/Be-9 ratio of these samples had been determined previously applying the conventional method where Be-10 and Be-9 concentrations are measured separately by AMS and ICP-MS, respectively. The resulting sedimentation rates are in discrepancy with values derived from biomarkers. To cross check the Be-10/Be-9 based age model two samples from each core were measured again with the new carrier-free method. The carrier-free results show systematically higher authigenic Be-10/Be-9 ratios. The calculated sedimentation rates of about 0.2 cm/kyr, however, are consistent for the carrier free and the conventional method. (C) 2012 Elsevier B.V. All rights reserved.