Many decisions involve uncertainty, yet most decision support systems (DSSs) and expert systems (ESs) are poorly equipped to deal with such problems. This paper employs an abductive perspective to propose a prototype framework for tackling uncertainty handling in DSS and ES. Five major features of such a framework, a user-friendly dialogue, a case-base, a knowledge-base, approximate reasoning and a fuzzification/defuzzification mechanism are presented and developed. These are supported by an extended example.