Good control of blood glucose concentration in insulin-dependent diabetics leads to fewer complications. Monitoring of glucose control is therefore central to the management of patients with diabetes. This is achieved both by the frequent measurement of blood glucose concentrations and by the measurement of blood glycated haemoglobin (sometimes referred to as haemoglobin A1c (HbA1c)). There have been a few isolated reports that glycated haemoglobin is subject to a seasonal variation in diabetics, both in children in Britain and in adults in Sweden. This is clearly important in interpreting changes in glycated haemoglobin in response to therapeutic interventions. Here we report that glycated haemoglobin in a British adult diabetic population is subject to seasonal variation. During a study investigating the effect of clinical glycated haemoglobin measurement on diabetic control (to be reported), it was noticed that there was an overall worsening of diabetic control (as assessed by glycated haemoglobin measurement) over the ®rst 6 months of the study (autumn 1995 to spring 1996), suggesting that there might be a systematic effect. Patients (both type 1 and type 2 diabetics) attending the diabetic clinic were therefore followed for a further 2 years. A total of 6721 results on 1295 mainly Caucasian patients [54% male, 51% type 1 diabetics aged 52+11 years, mean+standard deviation (SD)] were obtained using a dedicated cation exchange high-performance liquid chromatography (HPLC) system (the `Diamat’, Biorad Ltd, Hemel Hempstead, UK). Each patient provided an average of 5 2 samples (range 1±6, SD 2 3) at approximately 3-, 6or 12-month intervals over a 2-year period. The results (see Fig. 1) clearly show a seasonal variation, with a maximum in the spring (March, April, May) and a nadir in the autumn (September, October, November). The averages+SD for the autumn months of 1995, 1996 and 1997 during the study were 8 52+1 53 (n=783), 8 57+1 6 (n=695) and 8 51+1 6 (n=608). For the spring months of 1996 and 1997 they were 9 21+1 7 (n=623) and 9 04+1 7 (n=578). The differences between each autumn and each spring average glycated haemoglobin were all highly statistically signi®cant (P50 0001, Student’s unpaired two-tailed t-test). There is no reason to suppose that by chance more well controlled diabetics were seen in the spring than in the autumn, although if this were to be the case it would confound the results. The fact that two complete cycles of peaks and troughs were found also argues against this being a chance ®nding. During this period there was no change in the performance of the assay [the range of the monthly percentage coef®cients of variation (%CVs) for the low control at a glycated haemoglobin of 5 4% was 2 3±4 3, and for the high control at a glycated haemoglobin of 10 0% was 1 2±2 3]. These results are in broad agreement with previous studies in British children, which showed a peak in February and a trough in August, and in Swedish adults, which showed a peak in January and a trough in July, although the peaks and troughs observed in the present study occurred a little later. These results are most easily interpreted as there being a seasonal variation in blood glucose levels. It is interesting to note that the peak glycated haemoglobin found here was in the spring. Glycated haemoglobin is affected by the average of the blood glucose for about the previous 3 months. This would have extended into the previous winter months, the season of increased infections which would be expected to worsen diabetic control. Other seasonal factors, such as seasonal food availability, the amount of exercise undertaken or the seasonal changes in vitamin D levels, might also be important. These factors will necessarily vary from year to year, and so it is not surprising that the characteristics of the spring peaks are not identical. It has recently been shown in a cohort study of 21 healthy females that glycated Short Report Ann Clin Biochem 2001; 38: 59±60