With the acceleration of China’s urbanization process, the importance of energy conservation and emission reduction in the building sector has become increasingly prominent. The effective control of carbon emissions in coastal provinces has a decisive impact on achieving the carbon emissions peak target nationwide. Based on the analysis of the influencing factors, this study establishes an urban residential buildings carbon emission prediction model by combining the IPAT model and the ridge regression model. In addition, the prediction model is combined with scenario analysis to simulate the evolution of carbon emission trends of urban residential buildings in Fujian Province from 2018 to 2050 under different scenarios. The results show that total population, urban living area, residents’ consumption expenditure, urbanization rate, per capita GDP, and energy structure are key factors affecting carbon emissions from urban residential buildings in coastal cities. Only under the ultra-low carbon model scenario can Fujian’s urban residential buildings achieve the carbon peak goal in 2027 (13.4748 million tons of CO2), which requires a reduction of 59.67% compared to that under the baseline model scenario. This study can provide an effective reference for energy conservation and emission reduction work of the regional scale and even the national scale.