Ocean-bottom node acquisitions provide high-quality data but usually have large distances between the nodes because of cost. This makes the use of conventional processing difficult and has led to relatively little interest in such data for industrial purposes. We have considered a three-step workflow specifically designed for prestack depth imaging of P-waves recorded by ocean-bottom nodes. It consists of multiple attenuation, velocity model estimation, and prestack depth migration. Whereas multiple attenuation and tomography use data in the common-receiver domain, migration is performed in the common-angle domain. One of the main features is the handling of the sparse receiver geometry during velocity model estimation: the reciprocity of the PP-Green’s functions is used to obtain the required tomographic input using only the common-receiver gathers. The tomographic method also provides an estimate of the geologic dip, which is used to limit the migration operator. This provides migrated images free of migration smiles. The workflow contains no additional assumptions compared to those used to process ocean-bottom cable data. We validate the workflow on a 2D line extracted from a 3D real data set acquired in the North Sea. The results show that it is possible to use ocean-bottom data efficiently for prestack depth imaging.