On 11-12 June 2000, a small asymmetric mesoscale convective system (MCS) passed through eastern Colorado and western Kansas, producing severe winds and hail. The storm was monitored by two research polarimetric Doppler radars and a NEXRAD Doppler radar, as well as a VHF-based lightning mapping network and the National Lightning Detection Network (NLDN), as part of the Severe Thunderstorm Electrification and Precipitation Study (STEPS). The MCS produced severe-sized hail and severe winds throughout much of the ~5-hour period of study, from 2100 UTC on 11 June to 0150 UTC on 12 June. However, the electrical behavior of the storm varied significantly during this time period. Early on, around 2230 UTC, the storm peaked in both total (~600 min-1 for the entire system, according to the VHF lightning mapper) and cloud-to-ground lightning flash rate (over 15 min-1 at peak), and also featured significant positive CG lightning in both the convective and stratiform regions (approximately one-third of all CGs were positive around this time). The lightning mapper suggested most VHF emissions were occurring in the upper regions of the convective line, near 9 km MSL in altitude. Severe weather reports during this time period included up to 1.5-inch diameter hail and wind gusts up to 57 knots. However, after a period of decline there occurred a convective resurgence later in the lifetime of the MCS, around 0045 UTC. This resurgence also caused severe weather, with hail up to 0.88 inches in diameter. However, the electrical regime was very different from earlier, with negative CG flash rates varying between 5 and 10 min-1 during this period, and less than 5% of CGs were positive. In addition, though still intense (~300 min-1), total flash rates were not as high as the earlier peak, and VHF emissions were now centered around 5 km MSL in the convective line. In this study we plan to show the key kinematic and microphysical differences that led to the electrical differences between the two severe weather periods, and the implications of these results for using lightning data in the nowcasting of severe weather.