Background: Real-time estimated longevity has been reported in pacemakers for several years, and was recently introduced in implantable cardioverter-defibrillators (ICDs). Objective: We sought to evaluate the accuracy of this longevity estimate in St. Jude Medical (SJM) ICDs, especially as the device battery approaches depletion. Methods: Among patients with SJM ICDs who underwent generator replacements due to reaching elective replacement indicator (ERI) at our institution, we identified those with devices that provided longevity estimates and reviewed their device interrogations in the 18 months prior to ERI. Significant discrepancy was defined as a difference of more than 12 months between estimated and actual longevity at any point during this period. Results: Forty-six patients with Current/Promote devices formed the study group (40 cardiac resynchronization therapy [CRT] and 6 single/dual chamber). Of these, 34 (74%) had significant discrepancy between estimated and actual longevity (28 CRT and all single/dual). Longevity was significantly overestimated by the device algorithm (mean maximum discrepancy of 18.8 months), more in single/dual than CRT devices (30.5 vs. 17.1 months). Marked discrepancy was seen at voltages ≥2.57 volts, with maximum discrepancy at 2.57 volts (23 months). The overall longevity was higher in the discrepant group of CRT devices than in the nondiscrepant group (67 vs. 61 months, log-rank P = 0.03). Conclusions: There was significant overestimation of longevity in nearly three-fourths of Current/Promote SJM ICDs in the last 18 months prior to ERI. Longevity estimates of SJM ICDs may not be reliable for making clinical decisions on frequency of follow-up, as the battery approaches depletion.