We report 2-year data on effectiveness and tolerability of the GLP-1 analogue liraglutide in the EVIDENCE study. EVIDENCE is a 2-year multicentre, observational, post-marketing outpatient study evaluating liraglutide in clinical practice. Physicians in France recruited patients starting liraglutide treatment. Patients and physicians completed questionnaires at baseline, 3 months, 6 months, then at 6-month intervals. The primary objective was to determine the percentage of patients taking liraglutide and with HbA1c <7% after 2 years. We present data on effectiveness, tolerability and treatment persistence. Patient characteristics are described by mean (±SD) or frequency; baseline and study end measurements were compared by McNemar or Wilcoxon tests. Baseline data were from 3152 patients (53% male, age 59±11 years, BMI 34±7 kg/m2, diabetes duration 10±6 years, HbA1c 8.5±1.5%); 2029 patients (64.4%) were taking liraglutide at study end, beyond the objective of 1707. Ninety percent of patients (n=2804) had HbA1c ≥7% at baseline. The proportion with HbA1c <7% was higher after 2 years’ liraglutide treatment (n=759, 39.4%) vs. baseline (n=213, 11.0%; p<0.0001). Treatment significantly reduced HbA1c (-1.01±1.54%, p<0.0001), fasting plasma glucose (-0.32±0.63 g/L, p<0.0001) and weight (-4.09±6.97 kg, p<0.0001). Data are from patients with baseline and 2-year data available and still on liraglutide at study end (n=1928). Gastrointestinal disorders were reported by 261 patients (8.7%) and were the most common reason for withdrawal. In conclusion, these results show that 1-year findings were maintained at study end. Effectiveness of liraglutide in clinical practice is similar to that in randomised clinical trials (RCTs) (up to -1.5% HbA1c reductions and -3.24 kg weight loss). However, absence of a control makes it difficult to evaluate whether improvements are attributable to liraglutide alone. Incidence of gastrointestinal events was lower than in RCTs (<26.5%). These results suggest that liraglutide clinical trial data translate into therapeutic benefits in clinical practice.