The femur bone cancer pain model was developed by implanting mouse osteolytic tumor cells (NCTC 2472) into the intramedulla of the femur in C3H/HeN mice. In vivo imaging analysis revealed that the implanted tumor cells grew progressively over 14 days. Associated with the tumor growth, guarding behavior, which was an indication of ongoing pain, time-dependently increased. Limb use abnormality and allodynia, which were indications of ambulatory and neuropathic pain, respectively, also appeared. The analgesic effects of oxycodone and other opioids, such as morphine and fentanyl, were evaluated at 14 days when all pain-related behaviors clearly appeared. Oxycodone (2–20 mg/kg, s.c.), morphine (10–50 mg/kg, s.c.) and fentanyl (0.05–0.2 mg/kg, s.c.) significantly reduced guarding behavior. Oxycodone (5–20 mg/kg, s.c.) and fentanyl (0.1 and 0.2 mg/kg, s.c.) significantly reversed limb use abnormality, but morphine (5–50 mg/kg, s.c.) did not. Moreover, oxycodone (5–20 mg/kg, s.c.) dose-dependently reversed allodynia without affecting the sham-treated mice. Morphine (50 mg/kg, s.c.) and fentanyl (0.075–0.2 mg/kg, s.c.) also reversed allodynia, but morphine (50 mg/kg, s.c.) tended to affect and fentanyl (0.1 and 0.2 mg/kg, s.c.) affected the withdrawal threshold in sham-treated mice. These results suggested that oxycodone relieved not only ongoing pain, but also ambulatory and neuropathic pain, and that the analgesic profile of oxycodone could be different from that of either morphine or fentanyl.