AbstractEstablishing behavioral consistency in serial homicide offending behavior is essential when linking homicides together and to a common offender. An examination of 35 serial homicide behaviors utilized by 90 offenders in 450 serial homicide cases was carried out to identify whether these offenders consistently performed the same behaviors across their series of homicides and whether it is more effective to examine individual behaviors or groupings of behaviors in order to identify behavioral consistency. This study provides the first empirically based evidence testing the use of using single behaviors as linking factors in linking homicide cases in a series, and shows that this method has inherent difficulties in accomplishing this, thus providing the first wave of empirical testing that questions current theories in the literature on serial homicide, which has suggested that signatures or specific key behaviors are important in understanding the consistency in an offender's behavior across the crimes. Recommendations for future studies on linking using signatures are suggested, including a more in‐depth qualitative empirical evaluation of individual series. Copyright © 2007 John Wiley & Sons, Ltd.