Objective To evaluate the feasibility of virtual box block test ( VBBT) designed by force feedback device combined with virtual reality technology for remote evaluation of upper limb moter function. Methods A VBBT system for upper limb motor function assessment was designed. This system could be used to calculate the VBBT score, mean velocity, spectral arc length, energy consumption and mean grasping force of the subjects. In total 20 healthy subjects and 16 post-stroke patients were recruited to perform the VBBT. Results Significant differences were found in all VBBT parameters (P< 0. 05) between post-stroke patients and healthy subjects. Receiver operating characteristic ( ROC) curve analysis showed that the VBBT parameters werepossible to distinguish the two groups ( P< 0. 05). The correlations between VBBT parameters and traditional clinical assessments indicated the concurrent validity of the VBBT. Multiple regression analysis showed that the kinematic and kinetic parameters of the VBBT could partially explain the variance of the traditional clinical assessments. Conclusions The VBBT can be used as a remote assessment of upper limb motor function in post-stroke patients.