Previous studies have shown that across different modalities (i.e., audio, visual and audio-visual) males are less accurate than females when completing emotion recognition tasks. Although explanations for these gender-based behaviour patterns range from socio-cultural influences to psychological dispositions there is a lack of direct evidence why males tend to have more difficulties and are less accurate in recognizing emotions. In a pre-registered study (osf.io/w2tgr) of 312 males we aimed at filling this gap by examining whether (1) their ability in recognizing emotions is modality specific and (2) if variations in this ability are due to psychosocial (i.e., mood, empathy, personality, motives) or physiological factors (i.e., testosterone, cortisol). The results within the confines of this study informed our current understanding with regard to the audio-visual integration of emotional signals among men and enabled us to actively contribute to a scientific domain that is currently re-writing our understanding on the role psychosocial factors and hormones play in the recognition of emotions.