Background: Seasonal influenza infection is a major challenge in public health. Previous studies have shown that the timing of seasonal influenza epidemics varies with latitude, indicating that meteorological and environmental conditions are involved in the spread of influenza. We sought to investigate the link between the onset of influenza and meteorological factors in Gansu Province, to build an accurate forecasting model. Methods: We conducted time series analysis based on selected weather variables and influenza incidence data from 2006 to 2016 in Gansu Province. First, the cross-correlation function (CCF) was applied to explore the correlation between meteorological variables and influenza incidence. Then, a seasonal autoregressive integrated moving average model with additional covariates (SARIMAX) was performed to fit and predict influenza incidence. Results: After fitting the SARIMAX model, minimum temperature at lag 1 (β = −0.067, P < 0.05, 95% confidence interval: −0.119, −0.015) was negatively associated with loginfluenza incidence. SARIMAX (1, 1, 1) (0, 1, 1)12 with minimum temperature at lag 1 was the optimal model, with good prediction accuracy. Conclusions: Meteorological factors showed different effects on influenza incidence in Gansu Province. Our results verify the idea that climate is an important factor in the spread of influenza. Keywords: influenza; meteorological factor; time series; SARIMAX; Gansu Province