Cupping therapy has been used for the alleviation of muscle soreness in athletes. However, clinical studies of cupping therapy show conflicting results. Lack of standardized guidelines of the dose-response relationship of cupping therapy, such as appropriate cupping duration and negative pressure, limits the adoption of cupping therapy in clinical practice. The objectives of this study were to investigate the effect of various pressures and durations of cupping therapy on reducing muscle stiffness. The 2 × 2 factorial design with the repeated measures and counterbalanced design was used to test four cupping protocols, including two negative pressures at −225 and −300 mmHg and two durations at 5 and 10 min, in 12 healthy young people. B-mode and elastographic ultrasound was used to assess muscle stiffness of the triceps before and after cupping therapy. The region of interest of elastographic image was divided into the superficial and deep layers for assessing the effect of cupping therapy on stiffness of various depths of the triceps. Normalized stiffness was calculated as a ratio of pre-cupping stiffness divided by post-cupping stiffness of each participant. The two-way analysis of variance (ANOVA) was used to examine the main effects of the pressure and duration factors and the interaction effect between the pressure and duration factors. The results showed that there were no interactions between the pressure and duration factors (overall layer p = 0.149, superficial layer p = 0.632, and deep layer p = 0.491). The main effects of duration of the overall, superficial and deep layers were p = 0.538, p = 0.097 and p = 0.018, respectively. The results showed that 10-min cupping at -300 mmHg is more effective on reducing stiffness of the deep layer of the triceps compared to 5-min cupping (p = 0.031). This study provides the first evidence that the dose of cupping therapy could significantly affect changes of triceps stiffness and the deep layer of the muscle is more sensitive to cupping therapy compared to the superficial and overall layers.