Context. The Asteroid Redirect Mission (ARM) under development by NASA is being planned to collect a multi-meter boulder from a near-Earth asteroid (NEA), and to bring it to the cis-lunar space in the mid-2020’s for future study and exploitation by a crewed mission. The MarcoPolo-M5 project is being proposed in 2016 for the M5 mission opportunity by ESA, to bring back to Earth a sample from a very primitive D-type NEA. As D-types are very rare within the NEA population, considerable effort is still in progress to characterize easily accessible targets with unknown surface composition, in order to discover further asteroids that belong to this taxonomic group. Aims. We aim to further characterize the physical properties of two optimal targets for sample return space missions, the low-ΔV NEAs (341843) 2008 EV5 and (52381) 1993 HA. The asteroid 2008 EV5 is the baseline target of ARM, but only one spectrum of this object exists in the literature. The asteroid 1993 HA is a very favourable target for a space mission based on its dynamical properties: here we intend to assess if it is a suitable target for MarcoPolo-M5. Methods. We obtained visible spectroscopy of 2008 EV5 with the FORS2 instrument at ESO-VLT (Paranal, Chile), at different rotational phases. We also obtained visible and near-infrared spectroscopy of 1993 HA, using the EFOSC2 and SOfI instruments at ESO-NTT (La Silla, Chile). Visible photometry of 1993 HA was carried out within the IMPACTON project at the Observatório Astronômico do Sertão de Itaparica (Itacuruba, Brazil). Results. Our new observations are in agreement with the C-type classification of 2008 EV5, which is a requirement for the ARM mission. We obtained five visible spectra which do not show any variability within the limits of noise, suggesting a homogeneous surface. We obtained the first ever spectroscopic dataset (~0.4–1.6 μm) for 1993 HA, finding a featureless, red-sloped behaviour typical of D-types (a T or X classification is also possible, with decreasing confidence). We also found that the synodic rotation period of 1993 HA is 4.107 ± 0.002 h, a value that is optimal for the execution of a sample return mission. The derived lightcurve also suggests an elongated shape (axis ratio a/b ≥ 1.71). At this stage 1993 HA does indeed seem to be the most favourable target for MarcoPolo-M5, though future observations are necessary to study it further.