The invention relates to a distributed information hiding detection method facing mass web images, which comprises the following steps of: a, setting up a Hadoop operation environment on a cloud computing platform; b, generating a Nutch crawling queue; c, according to a MapReduce architecture of the Hadoop, carrying out distributed webpage crawling; d, analyzing a webpage and extracting images in the webpage; 3, storing the extracted web images by adopting Hbase; f, carrying out universal blind detection of the images; g, screening out images with a BMP (Bitmap) format and a JPEG (Joint Photographic Experts Group) format; h, detecting the images with the BMP format; and i, detecting the image with the JPEG format. The image hiding detection method disclosed by the invention adopts open source technologies such as Hadoop, Nutch and the like, is easy to design, develop and implement, has a good detection effect, is suitable for extraction of mass images and detection of hidden information in the current network environment and has important practical significance for finding out and tracking lawless people who utilize public channels to carry out covert communication and maintaining national information security.