At the previous NWP conference last summer in Ft. Lauderdale, the LAPS group of FSL presented a series of papers on the LAPS "hot start" technique for diabatic initialization of mesoscale models. Our verification results were computed over 40 wintertime cases for which the MM5 model was configured with a 10-km grid. The results were very encouraging, showing a distinct advantage in skill scores for 1-6 hr precipitation forecasts compared to identical model configurations using other methods of initialization. The International H2O Project (IHOP), which will be conducted during late May and June in Tornado Alley, gives us an ideal opportunity to validate the model against many different kinds of traditional and experimental observations, in a weather context focussed on convective meteorology. Our IHOP modeling configuration will have a fine mesh at 4 km over the domain of field operations, with a 12-km outer nest for lateral bounds, and we will run both the MM5 and WRF models, identically configured. This talk will present verification of these model runs and expose strengths and weaknesses of the hot start technique for convective weather.