To date, no microscopic methods are available to confirm scanning tunneling microscope (STM) images of DNA. The difficulties encountered in repeating these images may be attributed to inadequate distribution of molecules on the substrate, poor adhesion to the substrate, or the low conductivity of the molecules. However, these factors are difficult to assess in an STM experiment where they may act simultaneously. A method to isolate these factors involves partly masking the deposited molecules before coating them with a conductive film to produce adjacent segments of coated and bare DNA after the mask is removed. The coated DNA segments are conductive and mechanically stable to allow easy identifi'cation of DNA by the STM. Furthermore, the path of a molecule can be traced from a coated to an uncoated region to test STM imaging of bare DNA. Masked preparations of DNA deposited on platinum/carbon-coated mica and highly oriented pyrolytic graphite were examined with a tunneling current 1000 times lower than the usual nanoamps. The tip apparently displaces molecules adsorbed to graphite to preclude imaging whereas more stably bound DNA on platinum/carbon-coated mica appears in reversed contrast.Scanning tunneling micrographs have been reported for bare DNA on substrates in vacuum (1, 2), in air (3-14), and in solution (15). DNA bound to gold surfaces, either electrochemically (15) or by chemical modification of the substrate (14), can be imaged reproducibly, but the high-resolution views of DNA dried on untreated surfaces are difficult to repeat. In addition, artifacts have been reported on highly oriented pyrolytic graphite (HOPG), a commonly used substrate (16,17). As a result, whether or not the scanning tunneling microscope (STM) can image bare DNA at atomic resolution is uncertain. Nonetheless, the potential for determining the structures of biomolecules and their complexes in aqueous environments emphasizes the importance of developing STM and similar scanning probe microscopic techniques for biology. Factors that affect STM images of DNA are the distribution of molecules on the substrate, the strength of their attachment, and their conductivity. These factors were not isolated in previous experiments, so it is difficult to understand why high-resolution STM images of DNA are elusive. However, these parameters are separable using a method that creates contiguous segments of conductively coated and bare DNA by evaporating platinum/carbon (Pt/C) through a mask placed over the DNA molecules. Metal-covered DNA segments are conductive, stably attached to the surface, and easy to identify in an STM experiment, so the deposition can be confirmed (18). More importantly, scanning adjacent bare segments determines whether bare DNA can be imaged by the STM. Scanning tunneling micrographs of DNA adsorbed on HOPG and Pt/C-coated mica show that DNA is substantially less con-