Purpose: To assess the reproducibility of detection of osseous and extra-osseous metastases in cancer patients using whole-body diffusion-weighted imaging with background body signal suppression (WB-DWIBS). Material and methods: A prospective study was conducted on 39 consecutive patients (21 females, 18 males; mean age 48 years) with metastases, who underwent WB-DWIBS on a 1.5-T MR scanner. Image analysis was performed independently by two blinded observers. Inter-observer agreement was assessed for the detection of osseous (spinal, appendicular) and extra-osseous (hepatic, pulmonary, nodal, and peritoneal) metastases. Results: The overall inter-observer agreement of WB-DWIBS in the detection of osseous and extra-osseous metastases was excellent (k = 0.887, agreement = 94.44%, p = 0.001). There was excellent inter-observer agreement of both observers for the detection of osseous spinal (k = 0.846, agreement = 92.3%), osseous appendicular (k = 0.898, agreement = 94.8 %), hepatic (k = 0.847, agreement = 92.3%), pulmonary (k = 0.938, agreement = 97.4%), nodal metastases (k = 0.856, agreement = 94.9%), and peritoneal metastasis (k = 0.772, agreement = 94.9%). Conclusion: We concluded that WB-DWIBS is reproducible for detection of osseous and extra-osseous metastases in cancer patients.