Understanding the evolution of melt ponds on Arctic sea ice is important for climate model parameterisations, weather forecast models and process studies involving mass, energy and biogeochemical exchanges across the ocean-sea ice-atmosphere interface. A field campaign was conducted in a region of level first-year sea ice (FYI) in the central Canadian Arctic Archipelago (CAA), during the summer of 2012, to examine the potential for estimating melt pond fraction (fp) from satellite synthetic aperture radar (SAR). In this study, 5.5 GHz (C-band) dual co- (HH +VV - horizontal transmit and horizontal receive + vertical transmit and vertical receive) and cross-polarisation (HV + HH - horizontal transmit and vertical receive + horizontal transmit and horizontal receive) radar scatterometer measurements of melt-pond-covered FYI are combined with ice and pond properties to analyse the effects of in situ physical and morphological changes on backscatter parameters. Surface roughness statistics of ice and ponds are characterised and compared to the validity domains of the Bragg and integral equation model (IEM) scattering models. Experimental and model results are used to outline the potential and limitations of the co-polarisation ratio (VV/HH) for retrieving melt pond information, including fp, at large incidence angles (≥ 35°). Despite high variability in cross-polarisation ratio (HV/HH) magnitudes, increases at small incidence angles (> 30°) are attributed to the formation of ice lids on ponds. Implications of the results for pond information retrievals from satellite C-, L- and P-band SARs are discussed.