To explore meteorological and environmental impacts on human health, a distributed lag non-linear model and a generalized additive model were employed to study the exposure-response relationship between meteorological factors and respiratory system diseases from 2013 to 2016 in Funan, China The results showed that the decline in patient numbers occurs in the high temperature seasons. At a short time lag, high temperature is a risk factor, but high temperature reduces the number of patients with respiratory system disease at long time lags. Low relative humidity (RH) increases the risk of respiratory system diseases, and largescale 48-hour temperature changes (Delta T48) increase the risk of disease for a whole range of lags. There is a remarkable correlation between morbidity and meteorological factors, with temperature minima corresponding to peaks in the number of patients. For low RH, the lag effect is obvious for the first 4 days; for higher RH, the effect is weak. A back-propagation artificial neural network model constructed in this study was capable of effectively predicting respiratory system disease, using average temperature, Delta T48, RH, and maximum wind speed as inputs.