Background: In heart failure, patients with under-prescribing and under-dosing of angiotensin converting enzyme inhibitor (ACEI), angiotesinogen receptor blocker (ARB) or beta blockade (BB) had unfavorable consequences for the patients. But, in spite of the apparent clinical results, the translation of the results of such clinical results into clinical prescribing has been slow and incomplete in real practice. So, we evaluated under-prescribing and under-dosing of proven drug in acute systolic heart failure (SHF). Method: Between Oct 2003 and July 2009, 249 hospitalized patients with acute exacerbation (onset: 1 < month) of systolic heart failure were included. The blood BNP (Triage, Biosite) level, functional status and clinical parameters were analyzed on admission and at follow-up period (mean 3.8years). Bad prognostic parameter was defined as readmission by aggravated heart failure or cardiovascular death. Target dosages of ACE or ARB and BB were defined as AHA/ACC guideline. Results: 1. Mean age was 67.6짹13 and male was 47.0%. Most frequent etiology was valvular heart disease 28.3% (n=34). Mean ejection fraction of LV was 41.916.5%. NYHA III/IV class score was 58.8%. Readmission by clinical deterioration and death were 94 patients (37.8%). 2. Initial use of ACEI was 57.4% (n=143), ARB was 18.9%, BB was 54.8%, aldactone was 66.7%. Common combination of polypharmacy in initial use was ACEI or ARB + BB + Diuretics + Aldactone (n=111, 44.6%), ACEI + Diuretics + Aldactone (n=73, 29.3%), BB + Diuretics + Aldactone (n=25, 10.0%) and only diuretics (n=40, 16.1%). Maintain use of ACEI was 37.3% (n=93), ARB was 30.9%, BB was 49.8%, Aldactone was 56.2%. The patients with approaching target dosage of ACEI or ARB was 17.2%, BB was 17.7%. The patients with approaching target dosage of both drug was only 6.0% (n=15). 3. There were no significant differences for clinical parameters, baseline BNP level, and follow-up BNP level between patient with approaching target dose and not approaching dose except systolic blood pressure (147 vs. 135mmHg, p=0.026). But, in comparisons of under-prescribing, event rate was significantly higher in patients with not using of effective drug than others (p=0.011). By multiple logistic backward stepwise regression analysis, BNP reduction and usage of effective drug was significant predictive maker for bad prognosis (p=0.000, p=0.029). Conclusions: Our finding indicate most patients were used as initial effective drugs but, the only few patients were used target dosage for maintaining. Moreover, the under-using, but not under-dosing, of proven drug is significantly related with long-term outcomes.