BACKGROUND AbataCepT In rOutiNe clinical practice (ACTION; NCT02109666) was an observational study of patients with rheumatoid arthritis who initiated intravenous abatacept in clinical practice. We aimed to compare abatacept retention rates and clinical outcomes in patients from Germany versus other countries. METHODS Baseline characteristics, crude retention rates, and clinical outcomes were compared by treatment line in the German cohort at 2 years. In addition, biologic-naïve patients were compared with biologic-naïve patients pooled from other participating countries. RESULTS In the German cohort, 677/680 (99.6%) patients enrolled were evaluable and 171/677 (25.3%) were biologic naïve. At baseline, abatacept monotherapy was received by a similar proportion of biologic-naïve and biologic-failure patients in the German cohort, but by a greater proportion of biologic-naïve patients in German versus other countries cohort (27.5 vs. 12.9%). The overall crude abatacept retention rate at 2 years in the German cohort was 39.9%; retention rate did not differ significantly by treatment line, but among biologic-naïve patients it was lower in Germany than in the other countries cohort (42.1 vs. 58.7%; log-rank test p < 0.001). At 2 years, good/moderate European League Against Rheumatism (EULAR) response rates in biologic-naïve patients were 85.5% in the German and 92.1% in other countries cohort (p = 0.163). CONCLUSIONS In the German cohort of ACTION, abatacept retention at 2 years was similar in biologic-naïve and biologic-failure patients. Biologic-naïve patients in German cohort had a significantly lower abatacept retention rate and a trend of lower good/moderate EULAR response rate than those in the other countries cohort. CONCLUSIONS • Analyses of data from national patient cohorts provide insight on local treatment patterns. • In the German cohort of the ACTION study, abatacept retention at 2 years was similar in biologic-naïve and biologic-failure patients. • Biologic-naïve patients from the German cohort had a significantly lower abatacept retention rate and a trend of lower good/moderate EULAR response rate than patients from other countries. • Data from large international studies may not be directly applicable to individual countries.