The increase of endoscopic actions and infectious risks explains of interest in endoscopic material maintenance processing. This purpose study was to set to the point and validate a microbiological flexibles endoscopes channels (suction, biopsy, air/water) sample method. An inoculum made of faeces contents material and microorganisms suspension (Pseudomonas aeruginosa or Candida albicans), at low or high concentration, was injected into digestive endoscope's channels. For each microorganism, five trials were carried out at both of the two concentrations. After incubation, a global channels sample was performed with a recovering solution, then a counting was made. A variance analysis (ANOVA) concerning restated measures was performed for the whole results. The mean difference between the number of microorganisms injected and those recovered's is 0.114 log (IC95% [+0.088; +0.140]). It does not vary significantly according to the micro-organisms (p = 0.69) and the inoculum (p = 0.70). The coefficient of variation which is 0.525 (min: 0.010; max: 0.260) shows this method can be reproduced. This technique are allowed to be recommended to evaluate and to validate manuals and automatics endoscopes maintenance processing.