Image data for leaf venation networks collected from forest plots in Peru, supporting the CHAMBASA plant trait campaign. These data were originally archived to a separate database (ClearedLeavesDB) that has since become inactive. Images were obtained after chemical leaf clearing with a resolution of 179 pixels mm−1, with a full extent of 12.8 mm × 19.3 mm. Images (suffix '-CLAHE') were enhanced using in MATLAB by applying contrast limited adaptive histogram equalization to the green channel of each image, using a sliding window of 200 pixels and a gain of 0.01. For each image, a separate hand-segmentation is provided (suffix '-TRACED'), with the traceable region of interest in white and veins in red. Some portions of each image were not traceable due to sample preparation issues in the wet lab, or due to the preservation of the underlying leaf tissue. Raw unprocessed images are also provided for reference, but should only be used after contacting the authors. There may be filename standardization issues that are not resolved in this upload. Further information on sample naming and dataset provenance is provided in the two papers referenced below.