The purpose of this study was to investigate the dose-response of strength training for relieving musculoskeletal pain in female office workers. The relation between the dose of training in terms of total training volume (sets × repetitions × load reported in training diaries) during a 16-week strength training program and changes in pain (calculated as pain index, 0-100%, from self-reported intensities and durations of pain in the upper body and low back) was determined by regression analysis. The women were part of a randomized controlled trial with specific strength training (SRT), all-round physical exercise (APE), and a reference group (REF). Results showed that pain index in SRT and APE decreased significantly from baseline to follow-up (-25%/-22%) compared with changes in REF (-15%). In the dose-response analysis within the SRT group (n = 125), the total volume of training (mean 18.056 kg, SD = 13.798) was negatively correlated with changes in pain index (β = -0.16, p = 0.045), and there was a significant dose-response relationship between training volume per session and change in pain index (β = -0.20, p = 0.034). In contrast, training attendance (mean 1.69 sessions per week, SD = 0.8) was not significantly related to the change in pain index. In conclusion, achieving higher accumulated training volumes was important for reducing musculoskeletal pain in female office workers. The training volume per session should be optimized by securing a load at 10-15 repetition maximum and adhering to principles of progressive overload.