Fourier transform infrared spectroscopy by attenuated total reflection (ATR‐FTIR), combined with the partial least square (PLS) method provides a fast characterization of ethylene/butadiene copolymers' intricate composition. The PLS regression method is constructed to quantify ethylene, 1,2‐butadiene (vinyl), trans‐1,4‐butadiene, and 1,2‐cyclohexane units in the copolymer. These rings are formed by intramolecular cyclization during polymerization. The performance of PLS models is evaluated by comparing the result obtained by 13C NMR and the model for three unknown samples. It is shown that the proposed method allows to accurately estimate the chemical composition of ethylene/butadiene copolymers in a much shorter time than NMR.