Exercise has been proposed to increase serum testosterone concentrations. The analysis of existing literature demonstrates a large degree of variability in hormonal changes during exercise. In our manuscript, we summarized and reviewed the literature, and concluded that this variability can be explained by the effect of numerous factors, such as (a) the use of different types of exercise (e.g., endurance vs. resistance); (b) training intensity and/or duration of resting periods; (c) study populations (e.g., young vs. elderly; lean vs. obese; sedentary vs. athletes); and (d) the time point when serum testosterone was measured (e.g., during or immediately after vs. several minutes or hours after the exercise). Although exercise increases plasma testosterone concentrations, this effect depends on many factors, including the aforementioned ones. Future studies should focus on clarifying the metabolic and molecular mechanisms whereby exercise may affect serum testosterone concentrations in the short and long-terms, and furthermore, how this affects downstream mechanisms.