We present structure-oriented and edge-preserving algorithms for denoising seismic data volumes in the frequencyspace (fxy) domain. After transforming the 3D data to the fxy domain, we apply for each frequency slice, 1D edgepreserving smoothing filters which are guided spatially by the so-called gradient structure tensor. The synthesis of the denoised data shows that incoherent as well as footprint noise can be significantly reduced with no loss of important structural information. The algorithm can be efficiently implemented using linear convolutions to compute the structure tensor and simple 1D edge-preserving operators. Contrarily to other 2D and 3D edge preserving algorithms, the proposed filtering is not limited to impedance data volumes.Facultad de Ciencias Astronómicas y Geofísicas