Traditional logic and logic programming languages cannot handle uncertainty. Fuzzy logic can, but it requires a readily computable form. One possible way to achieve this is to define a propositional fuzzy logic, extend this to a first order form, convert it to Horn clause form, and, finally, to devise a theorem prover to manipulate the Horn clauses. The first two steps have already been published. The paper completes the exercise by deriving a fuzzy Horn clause logic from the first order logic already published, and proposing a theorem prover to manipulate these clauses.