Summary In Quebec, occupational deafness is the second most common disease, in terms of number of cases, to be compensated by the Commission des normes, de l’équité, de la santé et de la sécurité du travail (CNESST). Noise also heightens accident risk by masking warning signals, making it harder for workers to communicate and increasing worker fatigue. Reducing noise helps not only to cut the number of cases of occupational deafness and the associated costs, but also to limit this factor’s role in workplace accidents, while improving workers’ quality of life. Noise reduction in the workplace can be achieved by three means: reducing noise at source (less noisy machinery or enclosure of machinery), decreasing noise amplitude along its propagation path (installation of acoustical barriers or materials) or acting at the level of the worker (wearing hearing protection or limiting worker exposure time). The second means (acoustical materials) is the focus of this study. The effectiveness of a acoustical materials can be expressed in terms of an absorption coefficient, which is defined theoretically as a value between 0 (non-absorbent materials) and 1 (perfectly absorbent materials). Measurement of this coefficient, imprecise in a number of respects, can only be done in the lab. It is characterized by a large variability of results between testing laboratories, and the absorption values obtained often reach non-physical values in practice (greater than 1). Nonetheless, these values are currently used to select acoustical materials or to carry out estimated noise calculations (for which the main source of error remains the imprecise knowledge of these coefficients). On the basis of a first proof-of-concept, this study proposes a robust, reliable method for characterizing acoustical materials in the laboratory and assesses its applicability in the field, with a view to measuring the real performances of materials after they have been installed. The proposed approach is based on measuring the acoustic propagation between a mobile source (forming a virtual “antenna”) and a fixed dipole microphone located at a short distance from a sample of the material being tested. In a later post-treatment stage, the data are used to obtain absorption coefficients in diffuse sound field conditions. In conjunction with the study’s follow-up committee, five typical materials were chosen. These materials were tested in accordance with current standard methods: the reverberation chamber method and the impedance tube method. The relevant intrinsic physical parameters of each material were also measured for the purpose of modelling their acoustic absorption (Johnson-Champoux-Allard parameters: airflow resistivity, tortuosity, porosity, thermal characteristic length and viscous characteristic length). The result of this modelling was then used as a benchmark solution for the absorption coefficient value. The value obtained using the proposed approach was compared with this benchmark solution. Finite element simulations helped with proposing an optimum configuration for the dimensions of the tested material, and the geometry of the source antenna and the dipole microphone. The results obtained in the lab (in an anechoic chamber) and outside of the lab (in two ordinary rooms, in the presence of reverberation and background noise) confirmed the approach’s potential for measuring the realistic absorption of the various materials tested, generally consistent with the values obtained by numerical simulation. The approach does, however, sometimes produce an erroneous absorption value in the low-frequency range because of the propagation model used at the measurement post-treatment stage. An avenue for improving this model is proposed and partially studied.