Background: Diabetes is a common condition that often requires increasing intensity of glucose lowering regimens. We describe the population trends in the intensity of regimens, and associations of achieved HbA1c and treatment persistence. Methods: We performed an episode-based analysis of the EXTEND-45 dataset, assessing trends in glucose lowering therapy and the associated outcomes of HbA1c and treatment persistence. Trends from 2009 to 2014 were assessed for each intensity level of a glucose lowering therapy regimen, according to the year prescribed. Episodes were defined as the length of time that an individual adhered to a regimen through ongoing prescription, and this was used as to define persistence. Mean HbA1c were calculated for each episode. Persistence and HbA1c were compared across the different regimens of treatment intensity. Results: The intensity of glucose lowering therapy remained stable over time with around one third of episodes utilising a single glucose lowering agent. Mean HbA1c was higher for insulin-based treatment (mean 7.9 % SD = 1.3 %), and lowest for episodes of no glucose lowering treatment (mean 6.3 % (SD = 0.8 %). Around half of participants achieved glycemic targets of 7 %. While there was considerable variation in persistence, the median persistence was around 3 months (94 days, IQR 51–201 days). Conclusions: Therapeutic intensity for diabetes has remained stable over 9 years. Whilst there was considerable variability in persistence with glucose lowering regimens, the mean duration of all regimens was less than a year. Requirement for higher intensity treatment with insulin was related to poorer glycemic control.