BACKGROUND: We have previously observed dofetilide to be associated with improved survival when the pre-treatment baseline QTc interval was below 429 ms. In this study we tested the natural extension of this observation-that the same group of patients should have a loss of survival benefit after withdrawal of dofetilide. METHODS: Patients with congestive heart failure (CHF) and reduced left ventricular function enrolled in the Diamond-CHF (Danish Investigations of Arrhythmia and Mortality on Dofetilide-CHF) study were eligible for our QT substudy provided they were in sinus rhythm and had a measurable QTc interval from a 12-lead standard ECG taken before randomization to placebo or dofetilide. An extended follow-up was performed from study closure, December 1996 until August 2001. RESULTS: Of the 418 patients entering the extended follow-up, 215 (51%) patients died during this 4.5 years of additional observation time. The baseline QTc interval made no prognostic difference to mortality in placebo treated patients. For dofetilide treated patients with a baseline QTc interval <429 ms, increased mortality was observed during the extended follow-up compared to placebo (risk ratio 1.5, 95% confidence interval 1.1-2.8)[corrected]. CONCLUSIONS: This follow-up study shows significant loss of survival benefit upon withdrawal of dofetilide in patients with CHF and a pre-treatment QTc interval below 429 ms. An independent randomized trial is warranted to validate these results.