The trans-Atlantic dust transport has important implications for human and ecosystem health, the terrestrial and oceanic biogeochemical cycle, weather systems, and climate. This study provides an observation-based multiyear estimate of trans-Atlantic dust transport using a 7-year (2007–2013) record of the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) measurements of the three dimensional distribution of aerosol backscatter, extinction and depolarization ratio in both cloud-free and above-cloud conditions. We estimate that on a basis of the 7-year average and integration over 10°S–30°N, 182 Tg a− 1 dust leaves the coast of North Africa at 15°W, of which 132 Tg a− 1 and 43 Tg a− 1 reaches 35°W and 75°W, respectively. These flux estimates have an overall known uncertainty of ±(45–70)%. Because of lack of reliable observations, uncertainties associated with the diurnal variation of dust and the missing below-cloud dust cannot be quantified. Significant seasonal variations are observed in both the magnitude of total dust mass flux and its meridional and vertical distributions. The interannual variability of annual dust mass flux is highly anti-correlated with the prior-year Sahel Precipitation Index. Using only cloud-free aerosol observations to calculate dust mass flux could introduce a high bias when compared with all-sky conditions that include both cloud-free and above-cloud aerosol observations. The bias is about 20% at 35°W and 75°W in boreal winter and spring based on the 7-year average, as long as dust within and below low-level clouds is negligible.