This code performs an ensemble-based data assimilation for volcanic ash dispersion, combining the Ensemble Transform Kalman Filter (ETKF) method (Ott et al., 2004; Hunt et al., 2007) and the FALL3D ash dispersal model (Costa et al., 2006; Folch et al., 2009), named ETKF-FALL3D. This system produces a joint estimation of 3-D ash concentration and critical Eruptive Source Parameters (ESPs) that can improve the performance of the classical ash dispersion forecast strategies. This version of the code considers two uncertain ESPs, the column height and the Suzuki-A shape parameter, based on previous studies that shown that their uncertainty impact the most on the results. This version of the code has been designed to run with an Observation System Simulation Experiment (OSSE) where total ash mass load observations were simulated from a free model run. The data of the free model run and the generated observations are included. In the included free model run, the ESPs change with time. The provided scripts perform a 10 day data assimilation experiment which uses the simulated observations every 6 hours over a domain which is centered at Cordon Caulle volcano. The ETKF- FALL3D code is written in python and can be used with Python 2.7 or Python 3. FALL3D model is available through an open license (http://datasim.ov.ingv.it/models/fall3d.html), is written in FORTRAN-90 and is available for Unix/Linux/Mac X Operating Systems (OS).