Prediction of follow-up shock is crucial to stratify patients with dilated cardiomyopathy (DCM) requiring implantable cardioverter defibrillator (ICD). The objective of the article is to assess the predictive value of endo-myocardial biopsy (EMB) towards ICD shock and follow-up mortality. A series of patients with DCM scheduled for ICD implantation underwent EMB to further determine the genesis of DCM. Presence of fibrosis and inflammation was documented and related to outcomes. A total of 240 patients were referred for ICD as primary (56%) and secondary (44%) prophylaxis. EMB showed myocardial fibrosis in 55.4%, inflammation in 55.7%, and viral genomic material in 60%. Median follow-up was 39 months (1–209). Appropriate and inappropriate shocks occurred in 29.2 and 20.4%. At logistic regression, determinants of appropriate shock were ICD indication for secondary prophylaxis (direct relationship: p = 0.009, OR 3.4, CI 1.3–8.8) and presence of inflammation at EMB (inverse relationship: p = 0.04, OR 0.4, CI 0.1–0.9). Moreover, the sole determinant of inappropriate shock was age at implant (inverse relationship: p = 0.003, OR = 0.9, CI 0.90–0.98). Overall mean estimated survival was 168 months and 5-year survival was 83%. Degree of improvement in LVEF% was the sole determinant of follow-up mortality (inverse relationship p = 0.02; HR = 0.9; CI 0.88–0.99). Present selection criteria for ICDs implant rely mainly on LVEF% that lacks sensitivity and specificity. EMB can identify the substrate of increased or reduced life-threatening arrhythmias. Presence of inflammation is a positive prognostic factor for reduced arrhythmogenic risk, independently by the ICD implantation indication.