Cable forces play an important role in cable-stayed bridges and other cable-based structures. Any changes in its value will directly affect the structure's performance. Hence, the prediction of actual tension force in existing cables is essential. This research aims to study the comparison of two different placement of sensor configurations for three different cable end-support conditions in cable force prediction. Cable with hinged-hinged, hinged-fixed, and fixed–fixed end support are modeled. Random vibration is applied on the cable then vibration responses from two different configurations are captured and analyzed to obtain their frequencies and mode shapes. Effective vibration lengths are calculated by applying the curve fitting process to all detected mode shapes. Prediction of cable force is done using the formula derived from beam theory. The analysis figured out that the signal processing technique can successfully identify the frequencies and mode shapes for all cable models. Each cable support model has their own frequencies. Different sensor arrangements produce different mode shapes, but the effective length between the two cases is nearly the same. Only fixed–fixed-end cable models have quite significant differences. These effective length differences lead to relatively larger errors in tension prediction. Aside from that, the overall tension forces estimation shows promising results with an error of less than 1%.