The target of this research is to adapt a Neural Network Approach that could be used in determining proper vehicle speed upon relevant conditions data. These include weather condition such as: Precipitation intensity, speed of wind, and degree of visibility, and road surface condition. Data are input to neural network. Output driven is a decision about allowed (advised) speed of a vehicle along highway which is appropriate to the atmospheric and road surface conditions at locations of interest. Such application could reduce the possibility of accidents resulting from excess speed when bad weather condition.