BACKGROUND: The general increased life expectancy is reflected in the age of patients with inflammatory bowel disease (IBD). The knowledge about efficacy and safety of anti-tumour necrosis factor (TNF) therapy in elderly is scarce and conflicting. AIM: To assess the efficacy and safety of anti-TNF therapy in elderly patients taking into account eventual comorbidity. METHODS: Observational and retrospective single-centred study where 66 IBD patients initiating anti-TNF treatment at age ≥65 years (cases: ≥65 anti-TNF) were compared with 112 IBD patients initiating anti-TNF <65 years (controls <65 anti-TNF) and 61 anti-TNF naïve IBD patients treated with immunosuppressants (IMS) and/or corticosteroids (CS) ≥65 years (controls ≥65 IMS/CS). Controls were matched to cases for IBD type, follow-up, disease duration and anti-TNF type. Comorbidity was assessed by using the Charlson Comorbidity Index (CCI). Both efficacy and safety of treatment were adjusted for comorbidity. RESULTS: The short-term clinical response to anti-TNF at 10 weeks was significantly lower in cases: ≥65 anti-TNF (68% vs. 89%; P < 0.001), whereas at ≥6 months, differences were not significant (79.5% vs. 82.8%; P = 0.639). The risk for any severe adverse events was higher in cases: ≥65 anti-TNF than in controls <65 anti-TNF (RR = 4.7; P < 0.001) or controls ≥65 IMS/CS (RR = 3.09; P = 0.0008). Age older than 65 and CCI > 0 were independent risk factors for malignancy and mortality regardless of the medication. CONCLUSION: Elderly patients treated with anti-TNF have a lower rate of short-term clinical response and a higher rate of severe adverse events than the younger patients under the same treatment.