This study aimed at investigating the effects of different hand positions on the electromyographic (EMG) activity of shoulder muscles during the performance of the lat pull-down exercise. Ten healthy men performed 3 repetitions of the lat pull-down exercise using their experimentally determined 10RM (repetition maximum) weight. Four different common variations of the lat pull-down were used: close grip (CG), supinated grip (SG), wide grip anterior (WGA), and wide grip posterior (WGP). Normalized root mean square of the EMG (NrmsEMG) activity for the right posterior deltoid (PD), latissimus dorsi (LD), pectoralis major (PM), teres major (TM), and long head of the triceps (TLH) were recorded using surface electrodes and normalized using maximum voluntary contractions. Repeated measures analysis of variance for each muscle detected statistical differences (p < 0.05) in myoelectric activity among hand positions during both the concentric and eccentric phases of the exercise. During the concentric phase, NrmsEMG results for the LD included WGA > WGP, SG, CG. For the TLH: WGA > WGP, SG, CG and WGP > CG, SG. For the PD: CG, WGA, SG > WGP. For the PM: CG, WGA, SG > WGP. During the eccentric phase, the LD produced the following patterns: WGA > WGP, SG, CG and WGP > CG. The TLH pattern showed WGA > SG and CG. For the PD: CG > WGA, WGP. The results indicate that changes in handgrip position affect the activities of specific muscles during the lat pull-down movement. Also, performance of the lat pull-down exercise using the WGA hand position produces greater muscle activity in the LD than any other hand position during both the concentric or eccentric phases of the movement.