Five variations of TOM-CAST and two sources of weather data were used to schedule tomato early blight control for research trials at the Snyder Research and Extension Farm, Pittstown, N.J. TOM-CAST scheduled fungicide applications were initiated at 15, 25, or 35 disease severity values (DSV) and resprayed at 15 or 25 DSV. Weather data for generating the DSVs was obtained on-site with a Sensor Instruments Field Monitor™ or through subscription to the electronic meteorological service SkyBit, Inc. Bravo 720, 3 pints/acre, was used for disease control. Foliar disease, yields, and postharvest decays were evaluated. Daily DSVs, cumulative DSVs, and forecast spray schedule varied with weather data source. Because SkyBit data generated more DSVs during the season than Field Monitor data, the SkyBit-based forecasts called for one or two more sprays than the Field Monitor-based forecasts. However, the number of sprays actually applied was the same, one more or one less for each combination of initiation and respray thresholds. All treatment schedules reduced disease compared to the untreated control. Variation in initiation threshold did not affect disease control. All TOM-CAST schedules respraying at 15 to 20 DSV were as effective as the weekly schedule. All fungicide treatments increased total yields and reduced postharvest decays compared to the untreated control. Most treatments also increased marketable yields. The most efficient, effective Field Monitor-generated TOM-CAST schedule required nine sprays compared to 13 weekly sprays. The comparable SkyBit-generated schedule called for 10 applications. Chemical name used: tetrachloroisophtalonitrile (chlorothalonil).