Background Against the backdrop of global warming, extreme high-temperature weather events are becoming more frequent and pose a substantial threat to human health. Urumqi is located in a typical arid climate zone, but the health effects of heatwaves in this region remain insufficiently understood. Objective To evaluate the association between heatwave events and non-accidental mortality among residents in Urumqi, quantify the attributable mortality burden, identify vulnerable populations, and provide scientific evidence to support the development of targeted regional intervention strategies. Methods Data on non-accidental mortality, meteorological variables, and air pollutants in Urumqi from 2018 to 2024 were collected. A heatwave was defined as three or more consecutive days with daily mean temperature exceeding the 90th percentile of the study period. Under a time-stratified case-crossover design, a generalized additive model incorporating a cross-basis function within the distributed lag nonlinear model framework was applied to simultaneously assess the single-day (lag0-lag7) and cumulative (lag01-lag07) lag effects and exposure-response relationship of heatwave exposure with non-accidental mortality. Heatwave-attributable excess deaths was further estimated. Results The exposure-response relationship between daily mean temperature exposure and non-accidental mortality in Urumqi exhibited a J-shaped curve. Using 25 ℃ as the reference temperature, mortality risk increased when temperature exceeded 27.4 ℃. The cumulative relative mortality risk over lag07 following heatwave exposure was 1.100 (95%CI: 1.059, 1.142), with the highest effect observed at lag06 (RR=1.151, 95%CI: 1.095, 1.209). Subgroup analyses showed that females (lag06, RR=1.169, 95%CI: 1.082, 1.262) and residents aged ≥65 years (lag06, RR=1.156, 95%CI: 1.092, 1.224) had higher mortality risks than males (lag06, RR=1.140, 95%CI: 1.069, 1.216) and residents aged <65 years group (lag06, RR=1.141, 95%CI: 1.035, 1.257), respectively. The increased risk appeared earlier among females, beginning at lag02, compared with lag03 among males. Heatwave-attributable excess deaths showed an overall upward trend, peaking at 248 (95%CI: 152, 341) deaths in 2023 and 229 (95%CI: 141, 314) deaths in 2024. Conclusion Heatwave events are significantly associated with an increased risk of non-accidental mortality among residents in Urumqi, primarily through cumulative lag effects under consecutive days of exposure. Females and residents aged ≥65 years are identified as vulnerable populations. The estimated excess deaths attributable to heatwaves show an overall upward trend.