Coxiella burnetii, the agent of Q fever, is known to persist in humans and rodents but its cellular reservoir in hosts remains undetermined. We hypothesized that adipose tissue serves as a C. burnetii reservoir during bacterial latency. BALB/c and C57BL/6 mice were infected with C. burnetii by the intraperitoneal route or the intracheal route. Adipose tissue was tested for the presence of C. burnetii several months after infection. C. burnetii was detected in abdominal, inguinal and dorsal adipose tissue 4 months post-infection, when no bacteria were detected in blood, liver, lungs and spleen, regardless of the inoculation route and independently of mouse strain. The transfer of abdominal adipose tissue from convalescent BALB/c mice to naïve immunodeficient mice resulted in the infection of the recipient animals. It is likely that C. burnetii infects adipocytes in vivo because bacteria were found in adipocytes within adipose tissue and replicated within in vitro-differentiated adipocytes. In addition, C. burnetii induced a specific transcriptional program in in-vivo and in vitro-differentiated adipocytes, which was enriched in categories associated with inflammatory response, hormone response and cytoskeleton. These changes may account for bacterial replication in in-vitro and chronic infection in-vivo. Adipose tissue may be the reservoir in which C. burnetii persists for prolonged periods after apparent clinical cure. The mouse model of C. burnetii infection may be used to understand the relapses of Q fever and provide new perspectives to the follow-up of patients.