This dissertation deals with different issues of meta-programming for knowledge base systems. The typical elements on the meta-level of a knowledge base system, i.e. the assimilator and the query processor are considered. As a starting point a knowledge base system has been implemented using the meta-programming facilities of the logic programming language Godel and the ground representation of object-programs at the meta-level. The main emphasis was placed on the knowledge assimilation component of the knowledge base system; it contains update procedures that handle object-programs realised as (arbitrary) logic programs using the extended syntax provided by the Godel language. Two different ways of improving the efficiency of the meta-programs were investigated. The first one deals with integrity constraint simplification. Employing simplification methods has an important impact on the efficiency when checking whether a knowledge base satisfies its integrity constraint theory. The simplification method that has been implemented is the one proposed by Lloyd, Sonenberg and Topor. The second approach is an attempt to involve partial evaluation to specialise the update procedures of the assimilator with respect to a specific knowledge base and integrity constraint theory. It is known that partial evaluation can be used to improve the efficiency of programs, especially meta-programs which use the ground representation. Another aspect of this work was to assess the Godel meta-programming facilities and to give some experiences about how to use the language for declarative meta-programming with regard to knowledge base systems.