Downward continuation (DC) of gravity is an important step in airborne gravimetry, as it allows the reduction of gravity measurements taken at flight level to the Earth’s surface and/or the geoid. This study mainly evaluates and compares two different techniques for downward continuation: (1) the combination of Poisson integral, Fast Fourier Transform (FFT) and Gaussian filtering, and (2) the method of Least Squares Collocation (LSC) with a planar covariance function. The dataset used in this study is simulated data for a region of Central Macedonia in Greece coming from the International Centre for Global Earth Models (ICGEM) service. Meanwhile, data from real airborne gravimetry measurements obtained from the AirQuantumGrav 2023 campaign in Iceland, where data were collected using a strapdown airborne gravimeter, were utilized to generate modelled noise. DC is performed by employing gravity disturbances as input data after generating and applying different types of noise to the signal (White/Pink/Brownian/Modelled) and removing this noise by examining different filters such as Gaussian and Butterworth. The first technique involves the DC of filtered residuals of gravity disturbances, using the Poisson integral approach after symmetric data padding by 10% to mitigate edge effects. It was elaborated through a spectral approach with 1D-FFT and application of a Gaussian filter to reduce the noise added due to the DC approach. The second technique implements LSC by utilizing known auto-covariances and cross-covariances functions of the signal and noise. Both approaches employ the well-known Remove-Compute-Restore (RCR) technique to obtain residuals of gravity disturbances during the DC. Finally, the results at surface are evaluated by comparing truth data acquired from ICGEM. This work attempts to evaluate the ability of each technique to recover gravity disturbances on the topographic surface, while managing the challenges posed by the ill-posed nature of the DC problem.