This paper presents a new methodology to achieve the maximum adhesion between wheels of a train and rails throughout acceleration mode of electrical trains in terms of predicting its corresponding wheel slip. The proposed methodology obtains the best operation point holding the maximum acceleration at the maximum adhesion point. Since the feeble adhesion coefficient is the first problem of the adhesion control system and it is not a measurable quantity, an estimator is used to estimate the adhesion coefficient. Although variations of the adhesion coefficient versus the wheel slip do not have a fixed curve and a certain behavior, the proposed methodology is capable of gaining the maximum acceleration and optimizes the depreciation of the wheel and the rail. The interior loop of the proposed predictive slip control is the field-oriented control of the induction motor. After simulation, the suggested methodology is implemented on a test bench using a digital signal processor (DSP). The experimental and simulation results justify the efficiency and efficacy of the proposed idea.