Semantic Web Services frameworks like OWL-S and WSMO combine semantic descriptions of Web service capabilities, inputs, outputs and behavior with the syntactic interface descriptions in WSDL and XML Schema. The glue between the semantic and syntactic description layers is called grounding. In this paper we identify the uses for grounding and we present the existing grounding approaches and propose new ones, discussing their respective advantages and drawbacks. Finally, we compare OWLS and WSMO with regards to their support of the presented grounding approaches and we recommend which approaches should be considered for future work.