Nowadays stroke is becoming a leading cause of c.hronic disability in adults. Hand stroke patients loose the ability to control their hands and fingers. Rehabilitation of stroke patients with functional occupational therapy is conducted by training the patient to perform exercises from everyday life. A set of task oriented exercises utilizing haptic virtual environments was developed in our previous work; based on well established and common exercises known as Jebsen Test of Hand Function (JTHF). Quantizing the performance of the patient in an exercise is usually the responsibility of the therapist and is currently performed subjectively. The aim of this paper is to introduce a Fuzzy Inference System (FIS) that is able to utilize haptic data to quantize the patient's performance. However, when testing the system, it has been found that the CyberForce system is incapable of providing normative data for evaluating the patient performance due to calibration and consistency issues.