The nervous system is a target for physiological and protective effects of neuroactive steroids. Consequently, the assessment of their levels in nervous structures under physiological and pathological conditions is a top priority. To this aim, identification and quantification of pregnenolone (PREG), progesterone (PROG), dihydroprogesterone (DHP), tetrahydroprogesterone (THP), testosterone (T), dihydrotestosterone (DHT), 5α-androstan-3α, 17β-diol (3α-diol), 17α- and 17β-estradiol (17α-E and 17β-E) by liquid chromatography and tandem mass spectrometry (LC–MS/MS) has been set up. After validation, this method was applied to determine the levels of neuroactive steroids in central (i.e., cerebral cortex, cerebellum and spinal cord) and peripheral (i.e., brachial nerve) nervous system of control and diabetic rats. In controls only the brachial nerve had detectable levels of all these neuroactive steroids. In contrast, 17α-E in cerebellum, 17α-E, 17β-E, DHP and THP in cerebral cortex, and 17α-E, 17β-E and DHP in spinal cord were under the detection limit. Diabetes, induced by injection with streptozotocin, strongly affected the levels of some neuroactive steroids. In particular, the levels of PREG, PROG and T in cerebellum, of PROG, T and 3α-diol in cerebral cortex, of PROG, DHT and 3α-diol in spinal cord and of PREG, DHP, THP, T, DHT and 3α-diol in brachial nerve were significantly decreased. In conclusion, the data here reported demonstrate that the LC–MS/MS method allows the assessment of neuroactive steroids in the nervous system with high sensitivity and specificity and that diabetes strongly affects their levels, providing a further basis for new therapeutic tools based on neuroactive steroids aimed at counteracting diabetic neuropathy.