LITTLE ATTENTION HAS BEEN GIVEN TO SLEEPY NEAR- MISS DRIVING ACCIDENTS UP TO NOW DESPITE THEIR LIKELY RELATIONSHIP WITH ACTUAL DRIVING accidents. Although there have been investigations that included near-miss accidents and/or sleepy near-misses, these data were limited and not used to predict a sleepy accident.1-5 In fact, the focus of these and the following papers were mostly on sleep disorders and driv- ing accidents with near-misses an ancillary part of the data. These papers do not make clear any association between sleepy near- misses leading to an actual sleepy driving accident. Three investigations that have mentioned near-miss sleepy ac- cidents were reported in the sleep literature. Krieger et al.6 and Engleman et al.7 found that treatment of obstructive sleep apnea (OSA) with continuous positive airway pressure (CPAP) devices reduced the number of accidents and near-misses. Turkington et al8 used a questionnaire on driving history and a simulator to evaluate OSA performances. They reported that olderage female sex, and self-reported use of alcohol had the greatest influence on driving performance. However, self-reported near- miss accidents (in the previous 3 years) were independently as- sociated with poor performance (odds ratio 2.62, 95% CI 1.00 to 6.88). In addition, the subjective ESS score was independently associated with near-miss accidents (odds ratio 1.21, 95% CI 1.12 to 1.31). Further review of several sophisticated studies on sleepy driving reveals no data or specific focus on sleepy near-miss ac- cidents 9-13 Contrary to the minimal focus on sleepy driving near-misses in sleep research, industry has used this metric extensively both to predict accidents and to control near-miss precursors as a strategy to limit workplace accidents. Near-miss strategy has been effec- tively used in a broad spectrum of industries, including airlines, railroads, medicine, petrochemical processing, and nuclear pow- er.14-17 Use of a near-miss strategy system requires systematic re- porting of all near-misses that might be associated with a defined accident outcome. A near-miss may be defined as a detected event that has not caused any harm and therefore has limited immediate impact. The success of using this metric is predicated on sufficient education and knowledge of the incident to recognize and identify a precursor condition that may lead to a serious consequence. Near- miss events when accumulated and evaluated provide insight and early detection of a system’s possible weakness. Phimister et al.14 reviewed adverse incidents in industry and described the use of a Safety Pyramid where near-miss accidents fill most (two-thirds) of the lower pyramid and the actual adverse accident is at the pin- nacle. This pyramid can have a ratio as high as 300:1 industrial near-miss–no-harm accidents to accidents with harm.18 We are unaware of an adequately powered study with emphasis on near-miss accidents while driving sleepy due to sleep depriva-