Objective: Compare duration of mechanical ventilation (MV), weaning time, ICU-LOS (ICU-LOS), efficacy and safety of remifentanil-based regimen with conventional sedation and analgesia. Design: Centre randomised, open-label, crossover, 'real-life' study. Setting: 15 Dutch hospitals. Patients: Adult medical and post-surgical ICU patients with anticipated short-term (2-3 days) MV. Interventions: Patient cohorts were randomised to remifentanil-based regimen (n = 96) with propofol as required, for a maximum of 10 days, or to conventional regimens (n = 109) of propofol, midazolam or lorazepam combined with fentanyl or morphine. Measurements and main results: Outcomes were weaning time, duration of MV, ICU-LOS, sedation- and analgesia levels, intensivist/ICU nurse satisfaction, adverse events, mean arterial pressure, heart rate. Median duration of ventilation (MV) was 5.1 days with conventional treatment versus 3.9 days with remifentanil (NS). The remifentanil-based regimen reduced median weaning time by 18.9 h (P = 0.0001). Median ICU-LOS was 7.9 days versus 5.9 days, respectively (NS). However, the treatment effects on duration of MV and ICU stay were time-dependent: patients were almost twice as likely to be extubated (P = 0.018) and discharged from the ICU (P = 0.05) on day 1-3. Propofol doses were reduced by 20% (P = 0.05). Remifentanil also improved sedation-agitation scores (P < 0.0001) and intensivist/ICU nurse satisfaction (P < 0.0001). All other outcomes were comparable. Conclusions: In patients with an expected short-term duration of MV, remifentanil significantly improves sedation and agitation levels and reduces weaning time. This contributes to a shorter duration of MV and ICU-LOS.