The greenhouse environmental parameters can be used to establish greenhouse mirco-climate model, which can combine with disease model for early warning, with aim of ecological controlling diseases to reduce pesticide usage, and protecting greenhouse ecological environment to ensure the agricultural product quality safety. Greenhouse canopy leaf temperature and air relative humidity models were established using energy balance and moisture balance principle inside the greenhouse. The leaf temperature model considered radiation heat transfer between the greenhouse crops, wall, soil and cover, plus the heat exchange caused by indoor net radiation and crop transpiration. Furthermore, the water dynamic balance in the greenhouse including leaf transpiration, soil evaporation, cover and leaf water vapor condensation, was considered to develop a relative humidity model. The primary infection and latent period warning models for cucumber downy mildew (Pseudoperonospora cubensis) were validated using the results of the leaf temperature and relative humidity model, and then the estimated disease occurrence date of cucumber downy mildew was compared with actual disease occurrence date of field observation. Finally, the results were verified by the measured temperature and humidity data of September and October, 2014. The results showed that the root mean square deviations (RMSDs) of the measured and estimated leaf temperature were 0.016 and 0.024 ℃, and the RMSDs of the measured and estimated air relative humidity were 0.15% and 0.13%, respectively. Combining the result of estimated temperature and humidity models, a cucumber disease early warning system was established to forecast the date of disease occurrence, which met with the real date. Thus, this work could provide the microenvironment data for the early warning system of cucumber diseases in solar greenhouses.