Tomato spotted wilt virus (TSWV) is an important disease of tobacco, reported at significant levels since early 90's. Chemical control, specifically imidacloprid (IMD) and acibenzolar-S-methyl (ASM) applications in the greenhouse before plants are transplanted, is used to manage TSWV in the US. In fields with 25% or less TSWV incidence, IMD provides efficient control. Reduced plant growth in the field has been reported with ASM and therefore its use has not been widely adapted. This issue imposes a challenge for controlling TSWV. Therefore, knowledge of expected risk for a given season and TSWV progress in a field within a season could be useful for disease management. The effect of weather on TSWV loss was investigated with logistic regression analysis. 

TSWV loss reported between 1993 and 2007 was used from a total of 58 counties in North Carolina that represents 95% of the tobacco counties. The results demonstrated that the sum of monthly average air temperature from December to February had a positive and the total precipitation for the same months a negative effect on TSWV loss. This model was validated in Virginia and North and South Carolina between 2009 and 2011. Predicted risk was in the range of reported one from the tobacco counties in North Carolina and Virginia but overestimated risk in South Carolina. A modified model has been implemented in Georgia since 2010. In a second step, we quantified the progress of TSWV incidence in a field. When quantifying TSWV progress, Degree-Days (DD) and the field's TSWV history were significant explanatory variables. DD with four lower thresholds (i. E., 10.5, 18, 20, and 26 °C) were used. The models were validated with data collected in 2009 and 2010 in North Carolina. Three of the four lower thresholds DD had a good predictability. The models' applicability is currently under investigation with data collected in South Carolina.