Based on the daily mortality data of residents in Wuhan from 1 January 2004 to 31 December 2008, the meteorological indicators and atmospheric pollution index of the same period, the distributed lag nonlinear model (DLNM) is used to assess the effects of daily mean temperature on deaths (caused by cardiovascular and respiratory), adjusted for both secular，seasonal trends and other confounders.The results indicate that the effect of Wuhan average daily temperature on the two diseases has a Jshaped curve. The cold effect on cardiovascular and respiratory has a delay of l to 4 days and persists for 8 to 20 days. The hot effect appeared acute and the highest at the first day for both cardiovascular and respiratory，but lasts for 2 days and followed by mortality displacement. It can be seen that in Wuhan both low and high temperature increase the risk of cardiovascular and respiratory mortality with delayed effects. Cold effects seem to last longer than heat effects.