Background. Steroidogenesis is a complex enzymatic process in which cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) play an important role. Phosphodiesterase-5 inhibitors(PDE5i) increase cGMP, improving NO availability. Objective. To investigate whether long-term, chronic treatment with the PDE5i Vardenafil affects adrenal and testicular steroidogenesis in diabetic men, using liquid chromatography-mass spectrometry (LC-MS/MS). Design. A longitudinal, prospective, investigator-started, randomized, placebo-controlled, double-blind, clinical-trial was carried out. Setting and participants. 54 male patients affected by T2DM diagnosed within the last 5 years were enrolled. 26 and 28 patients were assigned to the verum and placebo-group, respectively. Interventions. The study consisted of an enrolment phase, a treatment phase (24 weeks) (Vardenafil/placebo 10 mg twice-daily), and a follow-up phase (24 weeks). Outcome measurements. Progesterone (P), 17-hydroxyprogesterone (17OHP), androstenedione (A), testosterone (T), dehydroepiandrosterone (DHEA), DHEA sulphate (DHEAS), corticosterone, 11-deoxycortisol and cortisol (C), were evaluated using LC-MS/MS. Results. No differences were seen in sex testicular steroids between study and control group. For the adrenal gland, steroids were considered according to the zona in which they are produced. Considering steroids produced in the zona fasciculata, no significant differences were seen in 11-deoxycortisol and C among visits, both in the study and in the control group. For the zona reticularis, DHEA significantly decreased during treatment only in the study group (p=0.007). At post-hoc test DHEA showed higher levels at visit 2 and 8 than in other visits. The DHEAS/DHEAS ratio significantly increased during treatment only in the verum group. Considering the adrenal zona glomerulosa, corticosterone significantly changed among visits both in the study and in the control group (p<0.001). At post-hoc test, in both groups, corticosterone was significantly higher at visit 2 (p=0.028), 8 (p=0.003) and 10 (p=0.044), i.e. in coincidence with the complete clinical and instrumental examination performed only at these visits according to the study protocol. Conclusions. This is the first double-blind, placebo-controlled clinical-trial in which steroidogenesis is extensively investigated by LC-MS/MS in T2DM men chronically treated with Vardenafil for 6 months, and followed-up for 6 months after therapy-withdrawal. Chronically administered Vardenafil reduces DHEA levels and increases DHEAS/DHEA ratio as possible consequences of modulation of steroidogenic enzymes by tissue changes in cGMP and/or cAMP availability. A possibly stress-related increase in corticosterone is suggested for the first time.