In experimental tumors, a steep dose-response curve has been demonstrated for chemosensitive malignancies. In clinical practice, increasing dose over the conventional dose results in higher responses. These observations have prompted efforts to use chemotherapy in higher doses. Initial empiricism lends itself to more rational use of high-dose chemotherapy. Strategies are based on alteration in drug schedules [e.g., high-dose cytosine arabinoside (Ara-C)], drug-to-drug interaction [e.g., high-dose methotrexate (MTX) with citrovorum rescue], and drug pharmacokinetics (e.g., regional chemotherapy). In general, increasing the dose of cytotoxic agents leads to increasing toxicity. Methods to circumvent unacceptable toxicity have been explored. Autologous bone marrow transplant is one such method that provides new avenues of therapy and has resulted in the emergence of a new spectrum of dose-limiting toxicity. Despite the improved response rate, high-dose chemotherapy has resulted in short response durations and no significant impact on the survival of cancer patients as yet. The greater benefit of such a modality may be in the consolidation of drug-sensitive tumors (e.g., lymphoma and small cell carcinoma) with minimal residual disease after minimal conventional therapy. We review various factors and concepts as they relate to the current use of high-dose chemotherapy. We conclude that drug resistance is multifactorial, and one variable that can be manipulated by clinicians is drug dose. The drug concentration appears crucial in enhancing cytotoxicity. All frontiers examining this approach in clinical practice deserve continued enthusiasm and cautious exploration while other factors are being fully understood and exploited.