The hydrologic community has recently focused substantial attention on ungauged or poorly gauged catchments, since hydrological prediction under these conditions is highly uncertain, but represents the majority of practical applications. Catchments in Greece are usually ungauged, due to resource constrains, whilst in gauged areas the period of record is often short for safe estimation of highly parameterised hydrological models. This paper is driven by the Prediction in Ungauged Basins initiative aiming to estimate catchment responses using readily available data, i.e. topographical, soil and land use information. The parsimonious rainfall-runoff model developed in a Geographical Information System (GIS) environment is based on the Clark's synthetic unit hydrograph technique to estimate the hydrological response. The method requires estimation of the time of concentration (based on the curve number), the storage attenuation coefficient and the time area histogram of the catchment. The models further compared with a data-based modelling approach using a Transfer Function and the simulated streamflow is analysed to investigate similarities as well as to better understand possible extensions of the resulting unit hydrograph. A case study using 10-minute observed data from the 15 km(2) Lykorema catchment, Attica, Greece, highlights the potential of the GIS-based model to predict, at least, the dynamic characteristics of the runoff response in ungauged or poorly gauged catchments.