Arterial hypertension and poor glycaemic control are central to the development of microalbuminuria in insulin-dependent diabetes mellitus (IDDM). Recent consensus has established sensitive criteria for their detection and treatment, although the proportion of patients who may benefit is unclear. Between 1988 and 1990, we measured urinary albumin to creatinine concentration ratio (A/C) in 3,636 adult out-patients with IDDM of more than 3 years duration, serum creatinine under 133 mumol/l and who were not undergoing antihypertensive treatment. A/C indicating microalbuminuria (> or = 2.38/2.96 mg/mmol, male/female) was found in 620 of 3,451 patients without proteinuria, and associated with hypertension (blood pressure > or = 140 and/or 90 mmHg; p = 0.0016; rate: 39.6%), independent of diabetes duration (p = 0.0082) and male gender (p = 0.0350; relative risk = 1.16; 95% confidence interval: 1.01-1.32). Hypertension was less common among those with normal A/C (27.5%, p < 0.0001) but was positively related with diabetes duration. Of the 1,015 patients with A/C > or = 2.0 mg/mmol 529 were reexamined. Glycated haemoglobin levels exceeded 3 SD above the mean of normal in 84.3% of the 198 microalbuminuric patients (AER = 20-200 micrograms/min), but were comparably poor (79.2%) in normoalbuminuria. Duration of diabetes was inversely related to glycated haemoglobin only in microalbuminuria (0.05 < p < 0.1). Intervention to lower blood pressure remains mainly restricted to those patients with long-term diabetes and slower development of kidney disease. Near-normalisation of glycaemia remains the priority for the majority of patients with microalbuminuria.