Public spaces in cold-region historic districts often contain a mismatch between spatial accessibility and winter environmental usability. Highly accessible streets may be exposed to strong winter winds, whereas sheltered spaces may remain underused because of weak network connectivity. Taking the Xinmin Street historic district in Changchun, China, as a case study, this study develops a coupled analytical framework that integrates space syntax, ENVI-met microclimate simulation, and GIS-based overlay analysis to identify winter public-space potential under heritage conservation constraints. Visual integration derived from Depthmap was used to represent configurational accessibility, while pedestrian-level wind speed at 1.5 m was used to characterize winter wind exposure under a representative clear winter-day scenario. The results show that high-integration areas are concentrated along the main north–south street, but several of these segments coincide with wind corridors and therefore show limited winter usability. Conversely, low-wind areas are mainly located on the leeward sides of buildings and within relatively enclosed spaces, although some lack sufficient pedestrian connectivity. Low-intervention optimization, including link completion, entrance adjustment, permeability enhancement, and winter-priority route organization, increased the average integration value of low-wind spaces from 7.41 to 8.82. The framework helps planners and conservation authorities identify sheltered but under-connected spaces and supports heritage-sensitive renewal strategies for improving winter public-space usability in cold-region historic districts.