Ecosystem condition refers to the quality of an ecosystem, both natural and anthropogenic, such as urban ecosystems. It is typically measured through abiotic and biotic variables, some serving as proxies for ecosystem processes or functions on which the supply of ecosystem services depends. While the linkage between ecosystem condition and ecosystem services is widely recognized, including by the UN Statistical Standard for Ecosystem Accounting (SEEA-EA), the relationship is complex, often non-linear or delayed, and varies by service. Generally, better condition is expected to enhance service supply, while overexploitation of services can lead to degradation of condition. This study critically reviews urban ecosystem service studies that incorporate ecosystem condition, examining how and to what extent it is integrated into assessment methods and models. For each service within each case study, we recorded key information, including: (1) assessment method; (2) condition variables and their classification (SEEA-EA, EBVs); (3) biodiversity level (for biotic variables); (4) spatial and temporal explicitness; (5) whether actual or potential service supply was assessed; and (6) consideration of sustainability. We reviewed 110 studies, with regulating services being the most frequently assessed category. Structural and compositional state variables were most commonly used, followed by chemical and physical variables, while functional variables were rare and largely limited to phenological traits. Only a minority of studies accounted for temporal dynamics or assessed actual service supply. Even fewer addressed sustainability aspects that could inform about overexploitation or unmet demand. Our findings show that, while ecosystem condition variables are increasingly considered in urban ecosystem service assessments, their integration remains partial. It is often limited to static, qualitative or semi-qualitative ecosystem service assessments and tends to overlook considerations related to the sustainability of service flows.