The goal of this study is to improve the ability of forecasters at NWSO Greenville-Spartanburg, SC to recognize subtle tornadic signatures in strong winter cyclone environments. To accomplish this task, a tornado that struck without warning near Easley (Pickers County), SC at 910 pm EST on 7 January 1998 is being investigated. The location is about 25 miles west-southwest of the KGSP WSR-88D. The event occurred during the passage of a line of convection in the warm sector of a strengthening extratropical cyclone. In spite of the absence of traditional tornado signatures on radar, the wind damage pattern discovered during a post-storm survey indicated a sequence of events frequently associated with the life cycle of a tornado. The tornado had a path length of 1.5 miles and a width of 50 to 100 yards. The most destructive phase of the tornado produced F2 darnage. Wind damage (FO) occurred in the inflow area of the storm prior to the point where the tornado circulation engaged the ground. Significant wind damage (F1) also occurred north of the tornado track in the rear flank downdraft region. The only other severe weather reports documented in the western Carolinas and extreme northeast Georgia were several cases of wind damage weakly linked to convection. Thus, an examination of the Pickens County storm is heing performed to determine if certain characteristics can be discovered that will enhance the likelihood of warning for similar isolated tornadoes in the future. Preliminary analysis of the event has revealed several significant facts: The tornado-producing thunderstorm evolved from a small, fast-moving convective system that caused significant wind damage 90 to 100 miles to the southwest; a persistent, well-defined mesocyclone did not exist in the 30 minutes prior to tornadogenesis; the depth of the weak mesocyclonic circulation decreased dramatically at about the same time the wind damage began; the tornado formed in a splitting cell just north of a bowing line segment in a weak line echo wave pattern; the tornado did not occur near the maximum reflectivity in the line; and the beginning of the damage path coincided with a rapid increase of terrain height. Based upon the initial analysis of WSR-88D velocity and reflectivity data, one concludes that strong mesocyclones are not required for tornado development in the warm sectors of deepening winter cyclones. Important environmental factors not resolved due to the scope of this study probably contributed to tornadogenesis. Nonetheless, the radar data derived from this investigation will enhance the ability of forecasters to rapidly diagnose the potential for short-lived, non- supercell tornadoes that occur in the Carolinas during the cold season.