Objective To analyze and summarize the experiences and problems in diagnosis of anthrax in outbreaks, and to provide a basis for anthrax prevention and control. Methods Lesion exudates were collected from suspected anthrax patients using slides and sterile cotton swabs and venous blood samples were also collected from suspected cases. Smear microscopy, direct culture, and broth enrichment were used to isolate and culture Bacillus anthracis. Real-time PCR was used to amplify the genes such as pagA in the virulence plasmid, rpoB in the chromosome, and cap in the capsule plasmid. Enzyme-linked immunosorbent assay (ELISA) was used to detect double serum anthrax antibodies. Results A total of 17 specimens were collected in the laboratory of the Center for Disease Control and Prevention of Tongliao, Inner Mongolia, China, in August 2018. The time window between disease outbreak and sample collection was 1 day at least and 16 days at most, with a mean of 6 days. In the 17 suspected patients, 15 (88.24%) received antibiotic treatment before sample collection. Bacillus anthracis was isolated from 1 specimen, yielding a positive rate of 5.88% (1/17). The real-time PCR analysis of all the specimens gave a positive rate of 41.18% (7/17). The results of ELISA showed a positive rate of 83.33% (10/12) for the protective antibody. Conclusion The optimal time to collect samples is the key to the early identification of outbreak. Real-time PCR, which is a fast and sensitive approach for diagnosis, is recommended for application in anthrax outbreak. Double serum specimens should be collected from all the suspected patients to increase the diagnosis rate.