Abstract Background Chemotherapy is essential for long-term survival of osteosarcoma patients. However, the impact of dosage and dosage intensity (DI) of chemotherapeutic agents on patients with high-grade osteosarcoma is largely unknown. Objective The object of this study was to evaluate the influence of these dosage-related variables on treatment outcomes in terms of event-free survival (EFS). Methods PubMed was searched for relevant English-language articles. Two reviewers extracted data independently. Sufficient data were presented to calculate the planned total dosage and DI for doxorubicin, cisplatin, ifosfamide, and methotrexate. Univariate analysis and partial regression analysis were performed to determine the association of 5-year EFS and the planned total dosage and DI of each drug. Results Seventeen studies comprising 23 trial arms met the inclusion criteria. The analysis recruited a total of 2257 patients. The study period ranged from 1976 to 2006. Using univariate analysis, the planned dosage and DI of methotrexate and ifosfamide correlated with better 5-year EFS (P = 0.001 for methotrexate dosage; P = 0.030 for ifosfamide dosage; P < 0.001 for methotrexate DI; and P = 0.033 for ifosfamide DI). There was a trend toward worse 5-year EFS with increase of doxorubicin DI (P = 0.055). Based on the partial regression analysis, the association of doxorubicin DI and ifosfamide dosage and DI with EFS became no longer statistically significant, and the planned total dosage and DI of methotrexate remained significantly correlated with better 5-year EFS (P = 0.001 and P = 0.004, respectively). Conclusions We observed a correlation of higher planned methotrexate total dosage and DI with better treatment outcomes in osteosarcoma patients. The present study showed that methotrexate dosage and DI were important predictors of the clinical outcomes and provided the rationale of high-dosage methotrexate in the context of multi-agent chemotherapy.