Abstract. Detailed measurements of radiation, atmospheric and aerosol properties were carried out in summer 2013 during the Aerosol Direct Radiative Impact on the regional climate in the MEDiterranean region (ADRIMED) campaign in the framework of the Chemistry-Aerosol Mediterranean Experiment (ChArMEx) experiment. This study focusses on the characterization of infrared (IR) optical properties and direct radiative effects of mineral dust, based on three vertical profiles of atmospheric and aerosol properties and IR broadband and narrowband radiation from airborne measurements, made in conjunction with radiosonde and ground-based observations at Lampedusa, in the central Mediterranean. Satellite IR spectra from the Infrared Atmospheric Sounder Interferometer (IASI) are also included in the analysis. The atmospheric and aerosol properties are used as input to a radiative transfer model, and various IR radiation parameters (upward and downward irradiance, nadir and zenith brightness temperature at different altitudes) are calculated and compared with observations. The model calculations are made for different sets of dust particle size distribution (PSD) and refractive index (RI), derived from observations and from the literature. The main results of the analysis are that the IR dust radiative forcing is non-negligible and strongly depends on PSD and RI. When calculations are made using the in situ measured size distribution, it is possible to identify the refractive index that produces the best match with observed IR irradiances and brightness temperatures (BTs). The most appropriate refractive indices correspond to those determined from independent measurements of mineral dust aerosols from the source regions (Tunisia, Algeria, Morocco) of dust transported over Lampedusa, suggesting that differences in the source properties should be taken into account. With the in situ size distribution and the most appropriate refractive index the estimated dust IR radiative forcing efficiency is +23.7 W m−2 at the surface, −7.9 W m−2 within the atmosphere, and +15.8 W m−2 at the top of the atmosphere. The use of column-integrated dust PSD from AERONET may also produce a good agreement with measured irradiances and BTs, but with significantly different values of the RI. This implies large differences, up to a factor of 2.5 at surface, in the estimated dust radiative forcing, and in the IR heating rate. This study shows that spectrally resolved measurements of BTs are important to better constrain the dust IR optical properties, and to obtain a reliable estimate of its radiative effects. Efforts should be directed at obtaining an improved description of the dust size distribution and its vertical distribution, as well as at including regionally resolved optical properties.