PurposeOpinions are often linked to emotions and stress. It is well established that testosterone and cortisol are useful biomarkers of stress and can predict human emotion and behavior. Therefore, the purpose of this work was to explore whether changes in testosterone and cortisol concentration map to shifts in opinions.MethodsWe present three studies with healthy men, where we (1) monitored diurnal changes in testosterone and cortisol alongside political opinion change using a descriptive longitudinal design, (2) assessed testosterone, cortisol and political opinion change in an experimental trial using a physical exercise intervention designed to decrease cortisol concentration and/or increase testosterone concentration, and (3) monitored testosterone and cortisol change with political opinion in a trial with psychosocial intervention designed to increase cortisol concentration/decrease testosterone concentration.ResultsTestosterone concentration and opinion changes were observed across the day (p < 0.05). Furthermore, opinions changed similarly to exercise-induced increases in testosterone (8.6–11.5%), but inversely to exercise-induced reductions in cortisol (p < 0.05). Opinion changes also occurred in similar fashion to testosterone changes following psychosocial intervention (-5.6% to -10.0%), and inversely to cortisol changes (p < 0.05). Finally, opinion change and testosterone change correlated at the within person level (r = 0.17 to 0.33; p < 0.05).ConclusionWe showed natural fluctuations in testosterone and cortisol covaried in parallel with shifts in opinion. Further, given the emergence of relationships for opinion change with hormone change, we contend that physiological stress response (i.e., testosterone and cortisol change) may predict, or even cause, opinion change.