The objective of this paper is to develop a decision-making approach for selecting candidates in soft-handover procedure in 3th or 4th generation mobile communication through grey relational analysis of the series similarity and approximation. The multi-generating and second grey relational analysis procedure is applied to select candidates in soft-handover procedure with considerations of the velocity and acceleration similarity of multi-generating data. The validation of computer simulation models illustrate how the approach can be applied in candidates selection in soft-handover, and obtain the best results of feasibility and effectiveness for user equipment (UE) in 3th or 4th generation mobile communications. Moreover, the approach could be easily applied to soft-handover procedure for the mobile communication systems. In this proposed approach is performed to select the candidate target cells by UE instead of eNodeB. It could provide a first solution to choose the candidate target cells through comparing multiple measured data for candidate-selecting with the target communication cell.