A survey suggests that many research misconduct incidents in the United States go unreported to the Office of Research Integrity. Sandra L. Titus, James A. Wells and Lawrence J. Rhoades say it's time to change that. A survey of US researchers suggests that scientific misconduct is greatly under-reported. The Office of Research Integrity was told of only 201 instances of likely misconduct relating to work funded by the Department of Health and Human Services in three years. Yet extrapolation from the survey predicts that over 2,300 observations of potential misconduct are made yearly. Sandra Titus, James Wells and Lawrence Rhoades argue that science can and should clean up its act, and recommend six strategies to that end.