To investigate the influence of hydrogen-plasma treatment on diamond(100) surfaces, heavily boron (B)-doped HPHT diamond crystals were mechanically and chemo-mechanically polished, and exposed to a microwaveassisted hydrogen plasma on a time scale of several minutes. The resulting surface morphology was analyzed on macroscopic scales by stylus profilometry (PFM) and on microscopic scales by STM and AFM. The polished samples have a roughness of typically 100 pm rms (PFM), with no obvious anisotropic structures at the surface. After exposure of the B-doped diamond(100) to the H-plasma, the roughness increases dramatically, and pronounced anisotropic structures appear, these being closely aligned with the crystallographic axis' and planes. An exposure for 3 minutes to the plasma leads to an increase of the roughness to 2-4 nm rms (STM), and a 'brick-wall' pattern appears, formed by weak cusps running along 110 . Very frequently, the cusps are replaced by 'negative' pyramids that are bordered by {11X} facets. After an exposure of an additional 5 minutes, the surface roughness of the B-doped samples increases further to 20-40 nm rms (STM), and frequently exhibits a regular pattern with structures at a characteristic length scale of about 100 nm. Those structures are aligned approximately with 110 and they are faceted with faces of approximately {XX1}. These results will be discussed in terms of strain relaxation, similar to the surface roughening observed on SiGe/Si and anisotropic etching.Many studies have been undertaken on CVD diamond films, towards applications that utilize the outstanding properties of diamond. So far, unmet demands are being put on the quality of those films, especially if one is aiming at utilizing the diamond films for semiconductor applications. Homoepitaxially grown single-crystalline diamond films are closest to meeting such requirements. Even though, nowadays, many research groups are able to grow high-quality diamond films homoepitaxially, various questions related to, for example, the nucleation and growth mechanism, the role of hydrogen in the bulk and at the surface, and the evolution of the morphology upon growth and etching of the diamond remain unanswered.The surface morphology of diamond substrates and films has been investigated by several groups. Since the substrate morphology is at least partly transferred to the film (see, e.g., [1]) and, for example, the resulting step density of a homoepitaxially grown film might influence its electronic properties at the surface [2], an optimized substrate preparation is required. Kühnle and Weis have developed a chemomechanical polishing procedure that reduces the roughness of diamond {100} to less than 100 pm rms [3]. Others have used hydrogen-plasma etching to reduce the surface roughness, resulting in large atomically flat terraces [4,5]. On the other hand, etching of the diamond(100) with atomic hydrogen at an elevated substrate temperature has led to {111} faceting of the surface, which has been interpreted as anisotropic etching of th...