To assess the frequency of user error in selecting the appropriate code for glucose meter strips during performance of self-monitoring of blood glucose.The glucose meter code and corresponding glucose strip code were examined in 335 different patients during routine clinic visits. All patients had received instructions in meter use from a clinic nurse, a diabetes educator, or a pharmacist. Of the 335 patients, 134 were excluded from analysis because of unavailability of the meter or strip container or unreadability of the code on the strip container. Age, sex, type of diabetes, treatment method, and most recent glycated hemoglobin (HbA1c) value (before the observation period) were compared for patients with the correct (matching) codes and those for whom the codes did not match.Of the 201 study patients, 26 had type 1 diabetes mellitus and 175 had type 2 diabetes. Overall, 107 patients were being treated with insulin, 16 by continuous subcutaneous insulin infusion. Matching glucose meter and strip codes were found in 169 patients (84%); 32 (16%) had incorrect (nonmatching) codes (P<0.0001). The mean HbA1c was higher in patients with incorrect codes (8.2% versus 7.7%), but the difference was not significant (P = 0.4688). Twelve percent of patients with type 1 diabetes and 17% of patients with type 2 diabetes had incorrect codes; this difference was not significant (P = 0.3674). Codes were incorrect in 12% of patients with type 2 diabetes treated with insulin, in comparison with 20% of those treated with orally administered agents (P = 0.2338). No significant difference was noted between the groups on the basis of age or sex.A considerable number of patients with diabetes fail to use glucose meters properly. Clinical decisions based on these data can result in adverse events. We found no significant difference in patients having the wrong code on glucose meters when compared for age,