The recent rapid growth in commercial complex facilities, the increasing complexity of urban spatial structures, and the widespread use of map-based navigation applications have caused changes in the way users perceive space and their movement behavior when navigating urban spaces. However, empirical studies examining age-based spatial cognition and the effects of digital wayfinding on the level of spatial perception are relatively limited, resulting in insufficient consideration of age-specific spatial cognition in urban planning. Therefore, the purpose of this study was to present fundamental data for improving user-centered urban space planning and wayfinding systems by analyzing the characteristics of spatial perception behavior by age among visitors to the Dongdaemun area in Seoul and the key factors affecting the level of spatial perception. The target study site was a space in the Dongdaemun district of Seoul where Kevin Lynch's elements of city image appeared in a complex manner, and face-to-face surveys were conducted with 200 visitors. The spatial cognition measurement items were reorganized through Exploratory Factor Analysis (EFA), and age group, frequency of visits, first-visit characteristics, and map application usage were set as the major explanatory variables. The main analysis was performed using Ordinary Least Squares (OLS) regression, while sequence logistic regression and random forest analyses were additionally performed to examine the robustness of the findings. In addition, analyses supporting the research, such as descriptive statistics and one-way analysis of variance (ANOVA), were conducted. The analysis showed that age was significantly associated with the level of spatial cognition (p < .001). Spatial cognition increased among individuals in their 20s through their 50s and then declined among those aged 60 years and older, indicating an inverted U-shaped relationship. Additionally, the frequency of visits and familiarity with first-time navigation showed a positive association with spatial cognition, whereas the use of map applications exhibited a negative association. EFA identified two dimensions of spatial cognition: facility/function-based cognition and physical structure-based spatial cognition. The explanatory power of the reconstructed spatial cognition model (R² = 0.353) was greater than that of the original single-scale model (R² = 0.136). However, between the two spatial cognition dimensions, the facility- and function-based cognition model was not statistically significant (R² = 0.021, p = .528). Therefore, the explanatory variables examined in this study were significantly associated only with the physical structure-based spatial cognition model. This study confirmed that age, visiting experience, wayfinding behavior, and digital wayfinding use showed significant relationships with the level of spatial cognition. In particular, it was suggested that user-centered urban planning strategies, including the development of age-friendly wayfinding systems, landmark-oriented spatial designs, and the integration of digital and non-digital guidance systems, are necessary. By empirically examining age-related spatial cognition in a complex urban environment, this study provided meaningful exploratory insights and a foundation for future research on spatial cognition in diverse urban settings. These findings have practical implications for age-friendly urban planning and wayfinding.