Objective To monitor the adult mosquito densities from 2007 to 2015, analyzing pattern of mosquitoes' species, to provide a basic reference for mosquito control measures. Methods The light trapping method was used to monitor adult mosquito density from April to December, 2007-2015. The autoregressive integrated moving average model (ARIMA) was used to predict adult mosquito density in 2016. Results A total of 94 515 mosquitoes were captured from 2007 to 2015, with a total density of 6.48 per hour. The predominant mosquito species in different habitats were significantly different, Culex tritaeniorhynchus accounted for 50.83% in livestock shed area. The predominant mosquito species in other various habitats was Cx. pipiens quinquefasciatus which accounted for more than 51.00%. The peak season of adult mosquito during 9 years was in June, the density of mosquito was high during June to September. ARIMA(1,0,0) (2,1,1)₁₂ was successfully established, residual sequence was tested by white-noise(P>0.05), and the R² was 0.61. Conclusion The community composition and seasonal density fluctuation of mosquitoes in Changsha city are acquired basically and the ARIMA(1,0,0) (2,1,1)₁₂ model is able to predict the adult mosquito density. It is suggested that we should make more effort to control mosquitoes in the countryside and take measures in April and May before the onset of the peak mosquito density.