The geology at Kottamiya, Rehab City and Zahraa–Madinat–Nasr to the East of Cairo (Egypt) is composed of low-velocity sediments on top of a rigid rock basement. Such sediments include the loose sands of the Gebel Ahmar formation, marl and shales of Maadi formation, in addition to sparse quaternary soil covers. Due to the contrast of the seismic impedance with the underlying bedrock, these soft sediments have the potential of considerably amplifying the ground motion during an earthquake. For the evaluation of site-specific seismic hazard, we computed the seismic site response in these areas by developing 1-D velocity models and derived average seismic velocities, including Vs30. To do that, we applied different active and passive source techniques, including the horizontal to vertical Fourier spectral ratio of ambient vibration recordings and multichannel analysis of artificially generated surface waves. A set of models representing the velocity structure of the site is then obtained by combined inversion of Rayleigh wave dispersion curves and ellipticity functions. While dispersion curves are used to constrain the uppermost low-velocity part of the soil profile, ellipticity helps in resolving the structure at the depth of the sediment–bedrock interface. From the retrieved velocity models, numerical ground-motion amplification is finally derived using 1-D SH-wave transfer function. We account for uncertainty in amplification by using a statistical model that accounts for the misfit of all the inverted velocity profiles. The study reveals that the different sites experience an important frequency-dependent amplification, with largest amplification occurring at the resonance frequencies of the sites. Amplification up to a factor of 5 is found, with some variability depending on the soil type (Vs30 ranges between 340 and 415 m s−2). Moreover, amplification is expected in the frequency range that is important for buildings (0.8–10 Hz), which is additional confirmation for the need of microzonation analysis of the area. The obtained results will be used for the development of a new seismic hazard model.