Abstract: In order to explore the design and construction of cucumber powdery mildew warning system in solar greenhouse, internet of things technology was used to conduct the real- time dynamic monitoring of the incidence of cucumber powdery mildew and cucumber growth environment in solar greenhouse. The growth environment included temperature and humidity of air and soil. Logistic regression model was used to construct cucumber powdery mildew warning model. The results showed that humidity characteristic variable (maximum air humidity) and temperature characteristic variable (maximum air temperature) had significant effects on the incidence probability of cucumber powdery mildew in solar greenhouse. And it was feasible to construct cucumber powdery mildew warning system in solar greenhouse with internet of things.