Some authors proposed to lower the present postglucose GH nadir cut-off (i.e. < 1 microg/l) to that obtained in healthy subjects to establish remission of acromegaly. The aim of the study was to correlate GH nadir with hormonal and metabolic parameters and to confirm the adequacy of the current limit to define disease remission. Retrospective study of 40 acromegalic patients cured by surgery, followed by radiotherapy when appropriate, studied at the time of disease remission (Phase 1) and re-evaluated after at least 3 years' follow-up (median 6.5 year, Phase 2). GH nadir was evaluated in 44 sex- and age-matched controls. Symptom score, pituitary function, neuroradiological imaging, metabolic parameters (BMI, glucose metabolism, insulin sensitivity, lipid profile, blood pressure). The upper limit of the 'normal' GH nadir was fixed at 0.26 microg/l (mean + 2SD of controls). In Phase 1, GH nadir was < 0.26 microg/l in 16 patients (Group A) and > 0.26 microg/l in 24 patients (Group B). Group B had only slightly higher IGF-1 SDS (0.4 +/- 1.0 vs.- 0.1 +/- 1.0, P = ns) and lower body mass index (BMI) than Group A (26.2 +/- 2.4 vs. 30.6 +/- 4.5 kg/m(2), P < 0.005). GH nadir positively correlated with IGF-1 (P < 0.05, r = 0.32) and negatively with BMI (P < 0.05, r = 0.42). In Phase 2, all patients had IGF-1 levels in the normal range and GH nadir < 1 microg/l, both parameters being even lower than those found at the time of remission. No patient had either clinical or neuroradiological evidence of disease recurrence. The current GH nadir limit is still adequate to define both short- and long-lasting remission of acromegaly, independently of the type of definitive treatment. Patients with the lowest GH nadir should probably be monitored long-term for adequacy of their GH secretion.