Bone is one of the most complex sample materials used for radiocarbon (14C) dating. The installation of the EnvironMICADAS AMS at HEKAL (department of ICER) in 2011 required the adoption of new sample preparation techniques for small bone samples. Since then, hundreds of procedural background and known-age bones have been processed using our modified Longin method (MLM) and dated along with unknown samples. Their results are used in this study to assess the reproducibility of our current bone preparation method and the real uncertainty of the final age result. In addition, using the background samples, which are included in each bone measurement batch, blank correction of the unknown samples could also be performed. The mean F14C value of our bone blanks is generally better than 0.005 (∼42,500 BP) alongside 0.0013 SD. Good reproducibility was confirmed by the results of the laboratory known-age bone as well, where the standard deviation of the mean is better than 0.0025. In addition, the results of the three bone samples used in an ultrafiltration (UF) test study did not show notable differences from the ones obtained by our current protocol in 1σ uncertainty range but more experiments will be performed in the near future.