Phenolic compounds (PC) are the main substances present in Euphorbia umbellata (Pax) Bruyns (janaúba) methanolic fraction (FM) of bark crude extract (CE). The FM exhibits antioxidant and anti-inflammatory activities1-2. However, no scientific study describes the development of a pharmaceutical product incorporated with this extract to be used as topic anti-inflammatory product. Chitosan has received big attention because it is a promising natural polymer that can be used in the biomedical field and pharmaceutical industries and can be combined with natural extracts to obtain different formulations3. Thus, the goal of this study is to propose the development of chitosan membranes incorporated with the FM and evaluates the release profile of the PC from this matrix. The membranes were obtained from a chitosan solution (1.5% in 1% acetic acid) under agitation for 60 min, and after this time, sodium benzoate, PEG 400 and Tween 80 were added and homogenized. 7.5 g of this mixture was poured into a plate (support) and placed in an oven at 55°C for 4 h (blank sample/membrane). For the membranes containing FM, different amounts of FM were added (50% - M50FM and 100% - M100FM) to the mixture and the procedure executed as proposed to the blank samples. To develop the release experiment, a quantity of 10 mL of distilled water was added to each plate and after different intervals (0, 1, 2, 3, 4, 6 and 8 h) the fluid was collected, and the same volume of water was replaced4. The aliquots collected were lyophilized and stored in the refrigerator until the use. The samples were solubilized in methanol, and the analysis was performed according to the Folin-Ciocalteau method using a UV-Vis spectrophotometer (715 nm). The released concentrations were determinated using calibration curves based on gallic acid standard5. The amount of total PC released from the chitosan matrix showed different release profile according the concentration of FM added to the membranes formulation. For the M50FM sample, it was observed a fast release in the first hour of the test (137.01±4.54mg/g) and controlled, reaching a plateau after 4 h. The membrane M100FM showed a different behavior, the amount of PC released was slower (first hour only 14.11±0.67 mg/g were released), however, it showed a controlled profile during all the time test, releasing 259.04±9.18 mg/g in the end of 8 h (a plateau was not observed during the test). These differences can be due to the interactions between the amine groups of chitosan and the carboxyl groups of the PC. The sample M50FM presents lesser chemical interaction, allowing a higher hydration of the chitosan, with consequently, more control of the release process. On the other hand, the sample M100FM presents more chemical interaction between chitosan and PC, causing reduction of the membrane hydration, reducing the amount of the compounds released in the beginning of the test. The high interaction also reduces the ability of the chitosan to control the PC release, which usually occurs after the polysaccharide hydration and swelling.