This study aims to analyze the effects of a training program based on practice variability on the speed and accuracy of the tennis forehand approach to the net shot. The study sample consisted of 35 players of both genders, 22 men and 13 women (age 44 ± 10.9 years, height 1.73 ± 0.8 cm, and weight 74.7 ± 8.4 kg). Players were randomly distributed into two groups (control group = 18 and experimental group = 17). Both training groups worked a total volume of 4 weeks, seven sessions, and 15 min per session of forehand approach shot. Control group performed traditional training while experimental group trained with variability using wristband weights. The data obtained showed a large Group × Time interaction in the accuracy of the forehand approach shot, F(1, 16) = 28.034, p < .001, η2 = .637. Only the experimental group increased significantly in the accuracy after the program (51.4%, effect size = 1.3, p < .001), while no changes were achieved regarding hitting speed (1.2%, effect size  = 0.12, p = .62). The control group did not improve in any of the tested variables. These results indicate that variability of training using wrist weights is a valid option to improve forehand approach shot accuracy in recreational-level players. Although stroke speed was not increased, this type of training may be interesting as accuracy and technical control is commonly the main goal of training at this level.