In our changing world, floods are a threat of increasing importance causing major fatalities and economic losses. Within this perspective, flood extent mapping is of great importance for both damage assessment and improving flood forecasts. While flood mapping through optical imagery is often hampered by the presence of clouds, Synthetic Aperture Radar (SAR) sensors are capable of sensing in all weather conditions during both day and night. Moreover, recently launched missions such as the Sentinel-1 and COSMO-SkyMed constellations provide improved temporal and spatial resolutions, thus even further increasing the potential of SAR for systematic flood mapping and monitoring. Due to their specular reflectance properties, open water surfaces typically appear dark and homogeneous on SAR images. Classification of these images is generally performed using a pixel-based approach. Frequently used algorithms include histogram thresholding, active contour models and pixel-based change detection methods. The use of object-based approaches for SAR-based flood mapping remains rare, although a couple of studies have worked with a segmentation step. However, pixel-based approaches suffer from quite some drawbacks. Especially thresholding typically results in classification products that still include a large number of dispersed misclassified pixels, thus requiring a post-processing step. Although computationally more expensive, active contour models mostly lead to higher accuracies, which indicates the importance of spatial context. An object-based approach allows taking into account this spatial context, as well as some other relevant properties such as object shape, proximity and homogeneity. Moreover, it is possible to include some additional information sources such as elevation data, land cover data and optical imagery. This study aims at further investigating the potential of OBIA for SAR-based flood mapping applications. A range of established pixel-based approaches will serve as a benchmark. Preliminary results demonstrate the benefit of both segmenting the image into objects as well as incorporating additional information sources.