The kinematic rupture of the L'Aquila earthquake on the Paganica fault is reassessed with a highly comprehensive set of data. This very complete dataset is also used to explore the possibility that secondary faults may be involved in the coseismic process. The main characteristics of the slip distribution of the Paganica normal fault found by previous studies are confirmed: an initial propagation around and up-dip of the hypocenter, associated with a high velocity rupture (> 3:5 km=s); a main asperity centered about 6-7 km southeast of the hypocenter, rupturing in the subsequent stage, with a maximum slip of nearly 1.5 m; a total rupture measuring about 15 km along strike and 12 km along dip; and an overall low average velocity rupture (< 2 km=s). The statement that the main features of the slip distribution are confirmed by different studies (based on distinct inversion schemes) is by itself an achievement, knowing that many past studies provided strongly discrepant slip maps for common events. Moreover, we focus specifically on the existence of slip at shallow depth and its relations with the other known active faults in the vicinity of the Paganica fault. We find that (1) as found in previous studies, shallow slip near the ground surface up-dip of the hypocenter is incompatible with the available Synthetic Aperture Radar interferograms and (2) the only secondary fault that is likely to have been activated during the mainshock is the San Gregorio synthetic fault located 2-3 km southwest of the Paganica fault.Online Material: Figures of waveform fitting using different velocity models, time versus distance distribution of slip from the single-fault model, variations in the geometry of the Paganica fault, geometry of complex faults models and geodetic data modeling, uncertainty on the slip value and the rupture time, effect of noise on resolution tests, quantification of the improvement on the GPS data fitting with the preferred two faults model with respect to the single-fault model, and a posteriori modeling of leveling data.