Ecological gradients along river banks of the Mura River influence forest species composition. On the basis of 58 relevés of floodplain forests along the Mura River, the classification of vegetation plots was performed with the Pc-Ord program. The diagnostic species combination for three clusters revealed after classification was calculated by fidelity measure (phi-coefficient) and presented in an analytic table. Average Ellenberg inidicator values, stream distances and relative elevations of the relevés were passively projected onto PCA to show ecological relationships among them. Correlations of plant functional type and stream distance gradients were calculated with regression analysis. For the main edifiers response curves to the stream distance gradients weremade. The classification of the Mura floodplain forests has revealed three ecologically interpretable vegetation types: Salicetum albae (most humid and nutrient-rich sites), Fraxino-Ulmetum allietosum ursini (ecologically intermediate sites), F.-U. quercetosum robori (the driest and the least nutrient-rich sites). Zonation of vegetation is connected to distance from the closest stream which influences species distribution through ecological gradients of moisture and nutrient. The proportion of therophytes is significantly negatively correlated with the distance from the closest stream and the proportion of neophytes is significantly negatively correlated with distance from the main stream.