HbA1 levels were determined by a rapid chromatographic column test in 15 healthy subjects (HS) and in 15 maturity-onset diabetics (MOD), fasting and 2 h after glucose ingestion (100 g for HS, 50 g for MOD). Chromatography was carried out both before and after 6 h of red cell incubation in saline at 37 degrees C. HbA1 in HS at 0 and 120 min of OGTT was not significantly different, either before (6.24 +/- 0.61% and 6.22 +/- 0.62%) or after red cell incubation in saline (5.85 +/- 0.61% and 5.87 +/- 0.55%). Red cell incubation in saline significantly reduced HbA1 levels both at 0 and 120 min (2p less than 0.001). HbA1 in MOD before red cell incubation in saline, was slightly but significantly higher at 120 min (8.61 +/- 1.03%) than at 0 min (8.39 +/- 1.01%): 2p less than 0.001. After incubation in saline, this difference was cancelled (7.86 +/- 0.85% at 0 min and 7.97 +/- 0.83% at 120 min: n.s.). Post-incubation levels were lower than pre-incubation ones both at 0 and 120 min (2p less than 0.001). The HbA1 increment observed in MOD is significantly correlated (p less than 0.01; r=0.64) to the blood glucose increment observed at the glycemic peak. We conclude that hemoglobin glycosylation may show rapid changes also in MOD, reflecting blood glucose changes, whereas in HS the physiological glycemic excursions are not wide enough to produce rapid HbA1 changes. Since labile and stable HbA1 co-elute in the rapid chromatographic methods, red cell incubation in saline for 6 h at 37 degrees C is recommended as a simple procedure which allows the measurement of the stable fraction alone, i.e. the real index of long-term glycemic control, independent of rapid glycemic fluctuations.