An earlier investigation has shown that TaCl5-graphite is relatively stable against humidity for a few days. Graphite intercalated flakes were stored in water and afterwards in air. An electron microprobe analysis gives evidence that destruction is a two-step process. First, tantalum chloride partly hydrolyses between the host layers. Electron probe microanalysis data make it likely that tantalum oxychloride is formed in this step. From this compound, tantalum oxide, the final product of hydrolysis, forms in a second step.