The daily number of patients with upper respiratory tract infection (URI) in 49 community health service centers in Luohu of Shenzhen from 2014 to 2018 and meteorological data in the same period were collected to analyze the changing characteristics of the number of patients with URI at different time scales. The distributed lag non-linear model (DLNM) and generalized linear model (GLM) were used to study the relationship between different meteorological factors and the number of patients with URI in Shenzhen. The results show that there was a significant seasonal change for the number of patients with URI in Luohu of Shenzhen. The peak periods of cases were March to April in spring, July in summer and December to next January in winter, corresponding to Pure Brightness, Lesser Heat and Lesser Cold of the 24 solar terms, respectively. The DLNM model showed that air temperature was the main impact factor, and its effect on the number of patients with URI presented mainly cold effect, the relative risk (RR) reached the peak after 4 days lag (RR was equal to 1.041, the 95% confidence interval was between 1.022 and 1.060). Women were more affected by cold effect than men, and middle-aged and elderly people were more affected by cold effect than children. Another important factors were thermal effect in summer and the variable temperature in spring. The influence of humidity presented mainly low humidity effect, and the relative risk reached the peak (RR was equal to 1.058, the 95% confidence interval was between 1.049 and 1.068) on the same day. Pressure and wind speed showed high pressure effect and strong wind effect, and RR reached the highest after 1 day lag. In conclusion, cold air activities in winter and spring and their cold effects such as low temperature, low humidity and strong wind were the key factors to induce URI, followed by the impact of continuous high temperature in summer, both of them should be focused on timely prevention.