The sedation needs of critically ill patients have been recognized as a core component of critical care and meeting these is vital to assist recovery and ensure humane treatment. There is growing evidence to suggest that sedation requirements are not always optimally managed. Sub-optimal sedation incorporates both under- and over-sedation and has been linked to both short-term (e.g. length of stay) and long-term (e.g. psychological recovery) outcomes. Various strategies have been proposed to improve sedation management and address aspects of assessment as well as delivery of sedation. To assess the effects of protocol-directed sedation management on the duration of mechanical ventilation and other relevant patient outcomes in mechanically ventilated intensive care unit (ICU) patients. We looked at various outcomes and examined the role of bias in order to examine the level of evidence for this intervention. We searched the Cochrane Central Register of Controlled trials (CENTRAL) (2013; Issue 11), MEDLINE (OvidSP) (1990 to November 2013), EMBASE (OvidSP) (1990 to November 2013), CINAHL (BIREME host) (1990 to November 2013), Database of Abstracts of Reviews of Effects (DARE) (1990 to November 2013), LILACS (1990 to November 2013), Current Controlled Trials and US National Institutes of Health Clinical Research Studies (1990 to November 2013), and reference lists of articles. We re-ran the search in October 2014. We will deal with any studies of interest when we update the review. We included randomized controlled trials (RCTs) conducted in adult ICUs comparing management with and without protocol-directed sedation. Two authors screened the titles and abstracts and then the full-text reports identified from our electronic search. We assessed seven domains of potential risk of bias for the included studies. We examined the clinical, methodological and statistical heterogeneity and used the random-effects model for meta-analysis where we considered it appropriate. We calculated the mean difference (MD) for duration of mechanical ventilation and risk ratio (RR) for mortality across studies, with 95% confidence intervals (CI). We identified two eligible studies with 633 participants. Both included studies compared the use of protocol-directed sedation, specifically protocols delivered by nurses, with usual care. We rated the risk of selection bias due to random sequence generation low for one study and unclear for one study. The risk of selection bias related to allocation concealment was low for both studies. We also assessed detection and attrition bias as low for both studies while we considered performance bias high due to the inability to blind participants and clinicians in both studies. Risk due to other sources of bias, such as potential for contamination between groups and reporting bias, was considered unclear. There was no clear evidence of differences in duration of mechanical ventilation (MD -5.74 hours, 95% CI -62.01 to 50.53, low quality evidence), ICU length of stay (MD -0.62 days, 95% CI -2.97 to 1.73) and hospital length of stay (MD -3.78 days, 95% CI -8.54 to 0.97) between people being managed with protocol-directed sedation versus usual care. Similarly, there was no clear evidence of difference in hospital mortality between the two groups (RR 0.96, 95% CI 0.71 to 1.31, low quality evidence). ICU mortality was only reported in one study preventing pooling of data. There was no clear evidence of difference in the incidence of tracheostomy (RR 0.77, 95% CI 0.31 to 1.89). The studies reported few adverse event outcomes; one study reported self extubation while the other study reported re-intubation; given this difference in outcomes, pooling of data was not possible. There was significant heterogeneity between studies for duration of mechanical ventilation (I(2) = 86%, P value = 0.008), ICU length of stay (I(2) = 82%, P value = 0.02) and incidence of tracheostomy (I(2) = 76%, P value = 0.04), with one study finding a reduction in duration of mechanical ventilation and incidence of tracheostomy and the other study finding no difference. There is currently insufficient evidence to evaluate the effectiveness of protocol-directed sedation. Results from the two RCTs were conflicting, resulting in the quality of the body of evidence as a whole being assessed as low. Further studies, taking into account contextual and clinician characteristics in different ICU environments, are necessary to inform future practice. Methodological strategies to reduce the risk of bias need to be considered in future studies.