Abstract Introduction Hypertrophic cardiomyopathy (HCM) is characterized by a heterogeneous clinical expression with increased risk of sudden cardiac death (SCD) from ventricular arrhythmias (VAs).Several studies have shown that patients with malignant arrhythmias have increased electrical dispersion and in homogeneity of intraventricular conduction. Strain by echocardiography is an excellent tool for assessing regional and global left ventricular (LV) function and mechanical dispersion reflects heterogeneous myocardial contraction. We aimed to explore the value of strain parameters in prediction of VAs in HCM with LV preserved systolic function. Methods Retrospective observational study including all patients with HCM and ICD implanted in setting of primary prevention in our centre. Patients with LVEF < 40% or coronary artery disease were excluded. LV GLS was defined as the average of peak longitudinal strains from a 16 LV segments model, obtained from three apical views. Time to peak strain was defined as the time from onset Q/R wave on ECG to peak negative longitudinal strain during the entire cardiac cycle. Mechanical dispersion was defined as the standard deviation of time to peak negative strain in 16 LV segments. Patients with VA (group1) and patients without VA (group2) were compared. Results The study population included 48 patients, 63.3% of male gender. A family history of HCM was present in 64 pts (43%). All patients were under anti arrhythmic therapy (BB in 95.6%, other anti-arrhythmic in 28.2%). VAs (sustained and non sustained) were documented in 27 (55%) patients. The study groups did not differ regarding to mean age (54 ± 12 vs 56 ± 12 years, p = 0.67), male gender (54% vs 56%, p = 0.87) and BB therapy (91% vs 96%,p = 0.07). Mean LVEF was 58% in group 1 and 61% in group 1, p = 0.56; a LVOT gradient >30mmHg was present in 52% of group 1 pts and 45% of group 2 pts, p = 0.06. Mean wall thickness was 22mm vs 18 mm, p = 0.03, respectively. GLS was significantly lower in group 1 (- 13.9 ±3.4 vs -16.1 ±3.5, p = 0.02), mechanic dispersion was significantly higher in group 1(81 ± 14 vs 60 ± 12ms, respectively, p = 0.01. Using a multivariate logistic regression model including variables included in SCD HCM risk score proposed by ESC mechanical dispersion (OR: 1.54 (1.03–8.7), p = 0,03) was a strong and independent risk predictor of VA. Using optimal cut-off values from ROC analyses, patients with mechanical dispersion >67 ms (AUC = 0.82, p= 0.02) had more VAs. Conclusions Mechanical dispersion and GLS may help to identify HCM patients with high risk of VAs and SCD. Abstract P1541 Figure 1