Patients with adrenal insufficiency (AI) require lifelong glucocorticoid (GC) replacement therapy. Cortisol measurement in saliva is increasingly being used: we assessed salivary cortisol rhythm in outpatients with AI, to provide new insights regarding the management of GC treatment. Nineteen AI outpatients collected six saliva samples from awakening (Fa , before taking the morning GC therapy), during the day (F1.5h , F6h before the afternoon GC dose, F8.5h , F12h ) until bedtime (Fb ). We measured daily cortisol exposure by calculating the area under the curve (AUCFa→Fb ). Saliva samples were collected at baseline and one year after GG dose reduction (by at least 5mg of hydrocortisone). Hydrocortisone equivalents decreased from median 25mg/d (baseline, interquartile range IQR 20-27.5) to 15mg/d (IQR 15-20, P<.01). As expected, we observed a reduction in both daily cortisol exposure (AUCFa→Fb 23982nmol·h/L, IQR 12635-45369, to 14689nmol·h/L, IQR 7168-25378, P<.001) and salivary cortisol levels at F6h (24.8nmol/L, IQR 20.1-35.7, to 21nmol/L, IQR 8.7-29.2, P<.05) and Fb (8.7nmol/L, IQR 3.4-20.2, to 3.7nmol/L, IQR 3.0-5.8, P<.05). None of the patients developed signs or symptoms consistent with AI after GC reduction. Median diastolic blood pressure (DPB) values fell from baseline to the end of follow-up (87.5mm Hg, IQR 80-90, to 80mm Hg, IQR 80-85, P<.05). The AUCFa→Fb of patients at baseline was above the reference value (90th percentile of controls) in 12 patients (60%); after the dosage reduction, 30% of patients normalized their daily cortisol exposure (AUCFa→Fb ). The reduction in GC treatment in patients with AI resulted in better control of daily cortisol rhythm, measured with salivary cortisol, and in an improvement of DPB. Further studies are needed to ascertain if salivary cortisol could be used as a biomarker to manage GC replacement therapy.