<p>In this study we have investigated the relationship between lightning and ice-hydrometeors for thunderstorms over norther India. we have implemented a hydrometeor identification (HID) algorithm on a C-band polarimetric Doppler Weather Radar (DWR) data to get information on the ice-hydrometeors and utilized lightning flash data of both cloud-to-ground (CG) and intracloud (IC) type lightnings, from Indian Lightning Location Network (ILLN). Seven prominent thunderstorm events, comprising a total of 385 radar volume scans, have been used in this study. The study reveals a good spatial association between graupel and lightning flashes. Time evolution of ice-hydrometeors and lightning flashes during thunderstorms are in tandem. It is revealed that, decrease in cloud-top brightness temperature (from INSAT-3DR satellite) precedes lightning occurrence by more than an hour. Event-wise correlations between ice hydrometeor related parameters and lightning flash rates show higher values for IC than CG flashes, but less correlations when all the data are pooled across events, signifying higher inter-storm variability in the relationship for IC. Graupel Count (GC), i.e., the number of radar grid points identified as graupel, showed the highest pooled correlation coefficient value of 0.87 for CG lightning. For IC lightning also, GC has the highest correlation coefficient value (0.74). The pooled correlations are higher for CG lightning for all ice-hydrometeor parameters. The findings support the non-inductive collision mechanism of thunderstorm charging. Lag-correlation analysis (up to 60&#xa0;min of lag) has been performed. Lag correlation coefficient values greater than 0.65 (for CG lightning at 30&#xa0;min of lag) in case of certain ice-hydrometeor parameters, strongly indicate that they are potential precursors for lightning and could be used for nowcasting of lightning.</p>