N=157 participants completed a set of 10 congruent and 10 incongruent moral dilemmas (cf. Conway & Gawronski, 2013). They also provided two saliva samples for hormone analyses. All lab sessions started between 13:30 and 17:30 hrs since circadian variation in testosterone levels is comparatively small in the afternoon. Session time had no effect on testosterone levels. Testosterone in the pooled saliva samples was determined using liquid chromatography-tandem mass spectrometry. Testosterone was not detectable in four female participants (3 COC users and 1 free cycling woman). Furthermore, during data preprocessing a total of seven outliers (1 men, 4 COC users, 2 free cycling women) had to be excluded. When analyzing moral decisions, we used the following three parameters (cf. Conway & Gawronski, 2013): (1) the proportion of decisions for outcome-based options in incongruent dilemmas (i.e., the traditional score), independent estimates of (2) deontology and (3) utilitarianism. Testosterone associations depended on sex and oral contraceptive use. In men, testosterone was negatively associated with deontology, while in free cycling women testosterone was positively associated with utilitarianism. Correlations with the traditional score were not significant. Furthermore, no correlations were found in COC users. There were no correlations in any of the sub-samples between testosterone and reaction times in congruent or incongruent dilemmas. Furthermore, we found significant differences between men and women in the traditional score and utilitarianism while the difference in deontology was only marginally significant. Differences in hormone status (men vs. free cycling women vs. COC users) were also significant in the traditional score and utilitarianism with the lowest inclination for utilitarian decisions in free cycling women and the highest in men while COC users fell in between. The effect on deontology was not significant. A preprint of the study can be found here: https://psyarxiv.com/7vy5u/