The purpose of this study is to gain understanding of the relationship between ejaculation and serum testosterone level in men. The serum testosterone concentrations of 28 volunteers were investigated daily during abstinence periods after ejaculation for two phases. The authors found that the fluctuations of testosterone levels from the 2nd to 5th day of abstinence were minimal. On the 7th day of abstinence, however, a clear peak of serum testosterone appeared, reaching 145.7% of the baseline ( P < 0.01). No regular fluctuation was observed following continuous abstinence after the peak. Ejaculation is the precondition and beginning of the special periodic serum testosterone level variations, which would not occur without ejaculation. The results showed that ejaculation-caused variations were characterized by a peak on the 7th day of abstinence; and that the effective time of an ejaculation is 7 days minimum. These data are the first to document the phenomenon of the periodic change in serum testosterone level; the correlation between ejaculation and periodic change in the serum testosterone level, and the pattern and characteristics of the periodic change.