Background: Nearly 15% of DNA tests for BRCA1/2 results in the identification of an unclassified variant (UV). In DNA diagnostic laboratories in The Netherlands, a 4-group classification system (class I to IV) is in use (Bell et al.). Aim of this study was to investigate whether the UVs in different classes showed a significant difference in their in silico characteristics and would justify current differences in protocols for counselling with respect to communication to the counselees. Methods: Missense UVs in BRCA1/2 identified between 2002 and 2010 (n = 88) were analyzed. In silico analysis of UVs was performed using SIFT– analysis Grantham score and AGVGD for the predicted severity of amino acid substitutions. Each UV was classified to one of the four classes. Results: More than half of the UVs (n = 50) were predicted to be tolerated using SIFT-analysis. Accordingly, all these variants are scored as neutral (C0) by AGVGD. Of the remaining 38 UVs not tolerated using SIFT-analysis, 19 were scored as C0 (neutral), 8 were scored C15–C25 (intermediate) and 11 were scored C35 or higher (likely to be pathogenic). Although class III UVs more frequently show in silico parameter outcomes that are suspicious for a pathogenic effect, the observed differences are not absolute. Seven UVs classified in class II had similar in silico profiles with 7 UVs in class III. Conclusion: This study showed that, in general, in silico analysis is consistently applied and proved to be able to discriminate between the different classes of UVs. However, additional analyses will be required to classify the UVs with more accuracy. In order to reduce psychological distress in families in which a UV is identified, we propose that communication of a UV should not primarily depend on its class, but also on the possibility to perform additional research in the family.