Most cellulolytic enzymes consist of distinct catalytic and cellulose-binding domains (CBDs). Similar domain structures are also found in enzymes degrading other insoluble carbohydrates such as raw starch and chitin. Such binding domains improve the binding and facilitate the activity of the catalytic domain on the insoluble but not on soluble substrates. Based on their amino acid sequence similarities, the CBDs have been divided into several different families. Structure determination and subsequent mutagenesis studies have revealed that CBDs rely on several aromatic amino acids for binding to the cellulose surfaces. The CBDs binding to crystalline cellulose have different topologies but share similar rigid backbone structures for correct positioning of the side chains required for the substrate recognition and binding. CBDs represent ideal affinity tags for specific immobilisation of various other proteins to cellulose. Furthermore, improved understanding and control of their action will be important for the improvement of the biotechnological value of cellulolytic enzymes.