Early compliance with the sepsis resuscitation bundle has been suggested to reduce mortality. However, few data are available about the impact of late compliance with the bundle on outcomes. The aim of this study was to assess whether the completion of the resuscitation bundle within the first 6 h after admission to the intensive care unit (ICU), but beyond the specific time limit of the various bundle interventions, is related to an improvement in survival. This prospective, cohort study included 753 patients recruited from September 2005 until August 2010 with severe sepsis/septic shock in the three medical-surgical ICUs of a tertiary academic medical center. We assessed the compliance with the different tasks included in the resuscitation bundle. Furthermore, we ascertained within the first 6 h after ICU admission the compliance with those tasks not carried out within their specific time limits; we have termed this variable "bundle improvement in the ICU." Septic shock was present in 88.1%. The overall in-hospital mortality was 31.6%. In 51.5%, there was bundle improvement in the ICU; this variable was associated with a lower risk of mortality (adjusted hazard ratio, 0.52 [95% confidence interval, 0.34-0.78]). That association was observed only when the time from severe sepsis onset to ICU admission was 6 h or less. Importantly, similar results were found after excluding all patients with severe sepsis (rapid responders) and those with refractory shock (nonresponders). The task with highest improvement was the achievement of central venous oxygen saturation 70% or greater in 39% of patients. Compliance with the resuscitation bundle even beyond the recommended time is associated with improvement in survival in patients with severe sepsis/septic shock.