We evaluated the long-term effects of self-monitoring of blood glucose (SMBG) on glycemic control in a large unselected group of insulin-dependent diabetic (IDD) children and adolescents (N = 282) treated at a diabetes clinic. Among those who had been taught SMBG techniques (N = 229) and reported frequency of use (N = 209), only 26% reported monitoring three or more times per day. HbA1 levels of patients who monitored their blood most frequently did not differ from those who monitored blood less frequently or those who monitored only urine. Likewise, HbA1 levels of patients who monitored with machines did not differ from Chemstrip bG users. Accuracy was assessed in a subsample of 100 randomly selected Chemstrip bG users by comparing their Chemstrip reading with a laboratory value. Fifty-eight percent of the readings were within 20% of the laboratory value. Accuracy did not relate to frequency of monitoring or to HbA1 levels. These data suggest that frequency and accuracy of SMBG are independent and that neither ensures good glycemic control.