Highway slipperiness is one of the most dangerous meteorological phenomena for traffic participants. Exclusion of glazed frost formation can be only achieved on the basis of slipperiness forecasting. Winter slipperiness forecasting is rather complicated problem due to a lot of influencing factors and their non- deterministic character. Various types of winter slipperiness and specific forecasting features of this phenomenon are thoroughly considered in the paper. Forecasting methods for some highways and their sections are mainly statistical at the present moment. They are based on processing of a great number of empiric data. The paper proposes an empiric model for forecasting temperature pavement according to ambient temperature, humidity, wind direction and presence of vegetation along a highway. The obtained model can be used to forecast slipperiness formation during winter highway operation.