The evaluation of man-machine interface for the cab of a combine harvester often involves some factors that are qualitative, lacks of available information and can only be assessed by human judgments, which inevitably involve various kinds of uncertainties and incompleteness. In view of this, an evaluation method combined the evidential reasoning approach with fuzzy membership degree was proposed to build the evaluation model for the man-machine interface of a combine harvester cab, which overcame the defects of handling information with uncertainty and incompleteness. Man-machine interface element can be assessed to a grade interval if it cannot be precisely assessed to only one grade. The introduction of utility function simplified the process of alternative ranking and decision. Applied example showed that this method could handle uncertain information effectively and the representation of results was much more coincident with people’s habit to deal with uncertain information than other methods.