Real-time solar wind speed measurements derived from measurements by the Solar Wind Electron, Proton, and Alpha Monitor (SWEPAM) aboard the Advanced Composition Explorer (ACE) satellite are critical to specifying and forecasting current space weather and its impacts. When solar energetic protons produce high background levels in SWEPAM, the real-time solar wind speed measurements can be corrupted due to errors in the onboard algorithm. We analyzed 14 years of ACE real-time solar wind data for the years 1998 through 2011 to determine how to identify the contaminated measurements and what proxy might be substituted for these corrupt data. We find that good criteria for flagging contaminated data are that (1) the measured solar wind speeds are below 305 km s–1, and (2) the >10 MeV ion fluxes measured by the ACE Solar Isotope Spectrometer (SIS) are above 180 pfu. We also compare several potential proxies for the contaminated solar wind and determine that the best proxy depends on the last valid measured wind speed, vinitial, and the duration of the contamination period. For the first 25 h, the best proxy is simply vinitial. At later times, if vinitial < 400 km s–1, then vinitial continues to be the best proxy, while if vinitial > 400 km s–1, the best proxy is a linear function of the Kp geomagnetic index.