Clinical guidelines and clinical pathways are accepted and proven instruments for quality assurance and process optimization in the healthcare domain. To derive clinical pathways from clinical guidelines, the imprecise, non-formalized abstract guidelines must be formalized. The transfer of evidence-based knowledge (clinical guidelines) to care processes (clinical pathways) is not straightforward due to different information contents and semantical constructs. A complex step within this formalization process is the mark-up step and annotation of the text passages to terminologies. The Unified Medical Language System (UMLS) provides a common reference terminology as well as the semantic link for combining the clinical pathways to patient-specific information. This paper proposes a semi-automated mark-up and UMLS annotation for clinical guidelines by using natural language processing techniques. The algorithm has been tested and evaluated using a German breast cancer guideline.