OBJECTIVES: To determine the prevalence and incidence of Lantus vs NPH and premix insulin use in diabetes including the treatment characteristics, comorbidity and resource use in a representative population in Canada. METHODS: Records from a longitudinal population-based database of more than 225,000 primary care patients in southwestern Ontario, Canada were analyzed between January 1 2008 to September 30 2010. Patients were considered to have diabetes if at least one of the following conditions was met: 1) physician diagnosed type 1 or 2 diabetes; 2) > 1 measurement of HbA1c greater than the recommended target; or 3) at least one prescription for a diabetes medication.  RESULTS: A total of 76,077 adult patients with representative data were included between 2008-2010.  Prevalence of T2DM was 7.9% and Type 1 diabetes was 2.9%. Patients on Lantus had less hypertension, nephropathy or Stage 5 kidney disease than NPH or Premix insulin patients (p<0.05).  Patients receiving Premix insulin tended to have more primary care visits, ER visits, hospitalizations and total referrals than Lantus.  More patients received new scripts for NPH than Lantus or Premix insulin during the study period.  The average dose of Lantus was 10.5-10.7 units, with a high rate of annual renewal (89.8-96.6%) for the same dose or any dose (93.8-98.7%).  There were very few dose switches or discontinuations for Lantus while NPH and Premix insulin were renewed less, underwent more dose switches and less discontinuations.  CONCLUSIONS: In a real-world setting the prevalence of diabetes was similar to nationally reported data. Patients receiving Lantus tended to have less hypertension, nephropathy or Stage 5 kidney disease, while those who took Premix insulin utilized more health services than Lantus or NPH.  Lantus scripts were renewed more often, had less dose changes or switches and less discontinuations than NPH or Premix insulin.