Endotracheal intubation in critically ill patients is associated with a high risk of complications that tend to increase with multiple attempts at laryngoscopy. In this pilot study, we compared direct laryngoscopy (DL) with video-laryngoscopy (VL) with regard to the number of attempts and other clinical parameters during endotracheal intubation of critically ill patients performed by novice providers.Patients were randomized to either VL or DL for endotracheal intubation. Exclusion criteria for the study included: requirement for immediate endotracheal intubation, cervical spine precautions, anticipated difficult intubation, oxygen saturation < 90%, or systolic blood pressure < 80 mmHg despite resuscitation. The providers, predominantly non-anesthesiology residents in their first three years of postgraduate training, received a one-hour teaching and mannequin session prior to performing the procedures.Forty patients, mean age 65 (standard deviation, 16) yr were randomized to VL (n = 20) or DL (n = 20). Sixty percent of the patients received endotracheal intubation for respiratory failure, and all patients received a neuromuscular blocker. Multiple attempts were required in 25/40 (63%) patients, and this did not differ with technique (P = 1.0) Video-laryngoscopy resulted in improved glottic visualization with 85% of patients having a Cormack-Lehane grade 1 view compared with 30% of patients in the DL group (P < 0.001). Total time-to-intubation for VL was 221 sec (interquartile range [IQR 103-291]) vs 156 sec [IQR 67-220] for DL (P = 0.15). Video-laryngoscopy resulted in a lower median SaO(2) (86%) during endotracheal intubation [IQR 75-93] compared with a median SaO(2) of 95% in the DL group [IQR 85-99] (P = 0.04).Video-laryngoscopy resulted in improved glottic visualization compared with DL; however, this did not translate into improved clinical outcomes. The trial was registered on ClinicalTrials.gov number, NCT00911755.