OBJECTIVES:   To assess the persistence of type-2 diabetic patients treated with basal supported oral therapy (BOT) with insulin glargine (GLA) compared to NPH insulin (NPH). METHODS: We performed a retrospective cohort study using claims data for ambulatory prescriptions within the German statutory health-insurance scheme, based on a representative sample of more than 80 % of German community pharmacies. Insulin-naive patients treated with oral antidiabetic drugs (OAD) who were additionally initiating therapy with GLA or NPH between January 2003 and December 2006 were included and followed up until December 2007. Persistence was defined as the duration of time from initiation of GLA or NPH until switching to intensified conventional insulin therapy (ICT). A switch to ICT was assumed whenever a short-acting insulin was prescribed for the first time followed by at least one prescription of a long-acting insulin within six months. Univariate and multivariate Cox proportional hazards models were used to compare both cohorts. RESULTS: In total, 97,998 patients (61,070 Glargine and 36,928 NPH) were included. Within the observation period, 23.5 % of GLA patients and 28.0 % of NPH patients switched to ICT. On average, these patients stayed 388 days on GLA and 313 days on NPH, respectively (p <0.001, log-rank test). The risk of switching to ICT was significantly higher for NPH patients compared to GLA patients (unadjusted hazard ratio [HR] 1.34 (99 % CI: 1.29-1.38)). After adjustment for predefined covariables i.e., type of physician (general practitioner vs. specialist), region, insurance status (member, family member, retired), health insurance company, comedication, number of OAD, dose of basal insulin, the risk for NPH patients remained significantly higher (adjusted HR: 1.22 (99 % CI: 1.18-1.27). CONCLUSIONS: Type 2 diabetic patients under basal supported oral therapy treated with insulin glargine stay significantly longer on initial therapy until they switch to ICT when compared to NPH.