Hospital-wide mortality rates are widely used as a marker of health care quality, but it remains unclear how to most accurately measure them. In this classic study, investigators provided standard hospital discharge data to four vendors who independently calculated hospital mortality rates. These methods produced varying rates of hospital mortality, and classification of hospitals as either higher or lower than average differed depending on the estimation method. These findings demonstrate that in-hospital mortality remains difficult to estimate and underscore the need for caution in considering it a marker of hospital care quality.