Objective: Patients receiving long-term glucocorticoid treatment are at risk of developing adrenal insufficiency during treatment. We investigated the prevalence of prednisolone-induced adrenal insufficiency in the particular clinical situation where patients receive ongoing low-dose (5 mg/day) prednisolone treatment, a dose by itself too low to cover glucocorticoid needs during stress. Design and methods: Cross-sectional study in 42 patients with rheumatoid arthritis (29 women, aged 36-86 years) treated with 5 mg prednisolone/day, who had received prednisolone for ≥6 months (median: 66, range: 6-444 months). Adrenal function was evaluated by a 250 μg Synacthen test performed after mean 48.7 h prednisolone pause. Local assay-specific cut-off for normal adrenal function was P-cortisol ≥420 nmol/L 30 min after Synacthen injection. Results: Overall, 20 of the 42 patients (48%, 95% CI: 33-62%) had an insufficient adrenal response to the Synacthen test. Including only patients who had not received concomitant treatment with any other glucocorticoid formulas within the last 3 months, 13 of 33 patients (39%, 95% CI: 25-56%) had an insufficient response. Adrenocorticotrophic hormone (ACTH) concentrations were generally low and anti-adrenal antibodies were negative indicating secondary adrenal insufficiency as the most likely diagnosis. There was no correlation between duration of treatment and 30 min P-cortisol (P = 0.62). Adrenal function did not depend on sex or seropositivity of rheumatoid arthritis. Conclusion: We demonstrate a high prevalence of adrenal insufficiency during ongoing low-dose prednisolone treatment. The results urge to increase focus on the condition to ensure identification and correct management of insufficient patients during stress and withdrawal. Strategies for adrenal function evaluation during ongoing low-dose glucocorticoid treatment need to be established.