The aroma types of cream cheese affect its commercial value and consumer acceptability. However, the types of volatile substances and sensory characteristics of cream cheese at different fermentation stages are still unclear. Therefore, in this study, headspace solid-phase microextraction gas chromatography-mass spectrometry (HS-SPME-GC-MS) and headspace gas chromatography-ion mobility spectrometry (HS-GC-IMS) were used to analyze the volatile substances in cream cheese fermentation. Orthogonal partial least squares discriminant analysis (OPLS-DA), odor activity value (OAV), relative odor activity value (ROAV) and variable projection importance (VIP) were used to identify the characteristic flavor substances in cream cheese fermentation. Finally, the relationship between key flavor substances and sensory characteristics was determined by partial least squares (PLS) analysis. A total of 34 and 36 volatile organic compounds were identified by HS-SPME-GC-MS and HS-GC-MS, respectively, and 14 characteristic flavor substances were found, based on VIP, ROAV and OAV models. Combined with sensory analysis and flavor substance changes, it was found that the cream cheese fermented for 15 d had the best flavor and taste. This study reveals the characteristics and contribution of volatile substances in cream cheese at different fermentation stages, which provides new insights into improving flavor and quality control.