- Clinical practice guidelines (CPGs) are increas-ingly common in clinical medicine for prescrib-ing a set of rules that a physician should fol-low. Recent interest is in accurate retrieval of CPGs at the point of care. Examples are the CPGs digital libraries National Guideline Clear-inghouse (NGC) or Vaidurya (DeGeL), which are organized along predefined concept hierar-chies, like MeSH and UMLS. In this case, both browsing and concept-based search can be ap-plied. Mandatory step in enabling both ways to CPGs retrieval is manual classification of CPGs along the concepts hierarchy. This task is ex-tremely time consuming. Supervised learning approaches, where a classifier is trained based on a meaningful set of labeled examples is not a sat-isfying solution, because usually too few or no CPGs are provided as training set for each class. In this paper we present how to apply the Tax-SOM model for multi-classification. TaxSOM is an unsupervised technique that supports the physician in the classification of CPGs along the concepts hierarchy, even when no labeled ex-amples are available. This model exploits lexi-cal and topological information on the hierarchy to elaborate a classification hypothesis for any given CPG. We argue that such a kind of unsu-pervised classification can support a physician to classify CPGs by recommending the most prob-able classes. An experimental evaluation on var-ious concept hierarchies with hundreds of CPGs and categories provides the empirical evidence of the proposed technique.