This study aimed to detect and understand remotely sensed urban wetland dynamics as a sensitive indicator of the combined effects of human disturbances and climate impacts in the course of global change. To address this objective, the study developed technical approaches to detect and interpret wetland changes across spatial scales in complex urban landscapes. Using a series of Satellite Pour l'Observation de la Terre (SPOT) images covering 1992-2010, the study was conducted in the Kansas City metropolitan area of the USA, which has experienced significant urban sprawl in recent decades. As a fine-tuning of the traditional supervised image classification, a knowledge-based classification algorithm was developed to identify fine-scale, hidden wetlands that cannot be appropriately detected based on their spectral differentiability. The analyses of wetland change were implemented at the metropolitan, watershed, and sub-watershed scales as well as being based on the size of surface water bodies in order to reveal real pictures of urban wetland change trends in relation to major driving factors. The results of the study indicated that the knowledge-based classification approach improved the detection capability and accuracy of urban wetlands by finetuning the traditional classification results. The cross-scale analysis of detected land covers revealed that wetland dynamics varied in trend and magnitude from metropolitan, watersheds, to sub-watershed scales. The study found that increased precipitation swelled wetlands, which inflated the findings of remotely sensed wetland cover and related trend interpretation. During an 18 year study period, human development activities in the study area resulted in a large increase in impervious surfaces, which was mainly at the expense of farmland/grassland areas and some small wetlands in all urban watersheds. In contrast, increased precipitation in the region swelled large wetlands in particular. This mixed picture of urban wetland dynamics, associated with the analysis of underlying driving factors, provides a new baseline for relevant urban planning, management, and research in a global change perspective.