As global climate change intensifies, the impact of ambient temperature on public health and healthcare systems has become increasingly significant. However, evidence on the association between temperature and health costs remains limited. This study investigates the association between ambient temperature and hospitalization costs (HCs) for cardiovascular diseases (CVD) in Tianshui, China, using data from 2016 to 2019. A distributed lag nonlinear model was employed to analyze the effects of daily average temperature on CVD-related HCs, self-payments (SPs), and Medicare payments (MPs), and to conduct cost attribution analyses across different temperature ranges. The results revealed a W-shaped relationship between temperature and HCs, SPs, and MPs. Compared to the reference temperature Minimum Hospitalization Temperature (MHT) of 0.3 °C, ambient temperature-attributable HCs totaled 7.47 million CNY over the study period, accounting for 3.65% (95% CI: 3.44-3.87%) of total costs, with an annual average of 1.87 million CNY. Low temperatures (below MHT) were associated with a higher risk of HCs, particularly at extreme low temperatures (1st percentile, -4 °C), which increased HCs by 35.4% (95% CI: 35.0-35.9%) compared to the reference temperature. However, the majority of HCs were attributable to high temperatures (above MHT), representing approximately 2.29% (95% CI: 2.15-2.44%) of total costs. Notably, extreme temperatures (both low and high) had a greater impact on SPs than on MPs, with 43.7% of SPs and 32.2% of MPs attributable to extreme temperatures. Additionally, men and younger individuals (age <
65 years) bore the majority of HCs. These findings underscore the escalating health and economic burdens of climate change, particularly in terms of CVD hospitalization. Proactive and targeted strategies can mitigate public health risks, reduce financial burdens, and enhance the sustainability of public health insurance systems.