Evaluation of sensitivity of hydrograph components and landscape metrics extracted from various spatial resolutions reveals relationships between landscape metrics and outflow properties with the lowest error. In the present study, 14 Digital Elevation Models (DEMs) with different resolutions derived from a vector and raster-radar source were analyzed to evaluate sensitivity of relationships between simulated hydrograph components and landscape metrics using multiple regression methods in the Galazchai Watershed, West Azerbaijan Province, Iran. To this end, DEM, slope, flow length and direction and time of concentration were developed in the vector and raster scales of 1:25000 and Advanced Land Observing Satellite-1 (ALOS) with different resolutions using ArcGIS 10.5 and ArcHydro software. Accordingly, 588 Direct Runoff Hydrographs (DRHs) were produced and clustered using Clark's Instantaneous Unit Hydrograph (IUH) model. The results showed that the best regression fits for flood volume and peak discharge were determined with Number of Disjunct Core Areas (NDCA) and Patch Density (PD) landscape metrics. Furthermore, the best regression fitted for time to peak and base time were observed with Core Area (CA) and Normalized Landscape Shape Index (NLSI) landscape metrics. In addition, the sensitivity analysis shows that the most sensitive spatial resolutions in modeling relationship between flood volume and peak discharge were determined at 5, 10, 20, and 30 m. The results revealed that hydrograph components in association with landscape metrics had high sensitivity to spatial resolutions, while the relationships between hydrological components and landscape metrics without considering the optimal spatial resolutions resulted in unacceptable results.