This paper proposes a simple approach to measure the elbow joint angle (EJA) using galvanic coupling system (GCS), directly; whereas, the traditional methods involved in either complex machine-learning task or arm movement models in which the consideration of model parameters are not accurate very often. First, a correlation between the EJA and GCS data has been established by defining a polynomial function based on a simple six-impedance model of human upper arm, where the EJA (theta) has been achieved by moving the forearm along the sagittal and transverse planes with different loads (empty hand, 1 and 2 kg). The coefficients of the polynomial are estimated based on the polynomial fit technique in which the actual angles (reference frame) are calculated by using motion data. In total, eleven subjects (seven males and four females) with the age of 30 +/- 6 years have been considered during the experiment. However, the GCS data of eight subjects are used to derive the correlation, exclusively. Furthermore, the influence of muscle fatigue and different loads on the derived correlation has been studied. Next, based on the derived correlation, the EJA has been measured in two parts-inside and outside tests by considering six subjects. The results show that the proposed idea can measure the EJA very effectively with error up to +/- 0.11 rad (6 degrees). Moreover, in a performance comparison, the proposed approach shows its compatibility by indicating low complexity, higher accuracy, and easy to measure.