We processed 41 Voyager 2 images of Neptune’s moon Triton with pixel scales < 2 km/pixel form their raw, compressed, archived state to more usable cloud-optimized geotiffs, which can easily be used within spatial analysis software such as GIS. Processing was done using the USGS’ ISIS software and included geometric and radiometric calibration, and the removal of image reseaux and corner markers (originally used for geometric calibration). The images were also photogrammetrically controlled relative to one another to improve their locations on the surface, which were initially inaccurate by up to 200 km. After performing a least squares bundle adjustment, the root mean square (RMS) uncertainty in image locations was 0.50, 0.52, and 0.51 pixels in latitude, longitude, and radius, respectively, with minimum and maximum residuals of -4.21 and +3.197 pixels, respectively. Each individual image was warped to an orthographic projection centered at 15o W and 18o N at the native image resolution. Because reseaux removal introduces interpolated data, two versions of each image are provided: a fully processed version with reseaux removed, and a partially processed version that retains the reseaux (i.e., no interpolation). We also generated a mosaic of Triton images that is spatially consistent with the entire dataset and provides context for the individual images (not every image is included in the mosaic). The mosaic uses the same orthographic projection as the individual images, but a consistent scale of 600 m/pixel. Three versions of the mosaic are included: a fully processed version (reseaux removed, some interpolated pixel values), a partially processed version (reseaux retained, no interpolation), and a fully processed but sharpened (high pass filter with 100% albedo add-back) to enhance surface features. This data release improves the usability and accessibility of this singular dataset and enables new scientific investigations of Triton.