Monoclonal antibodies are increasingly being used as protein therapeutics for cancer. They offer very specific binding to target molecules on the surface of cancer cells, relatively few side effects and predictable pharmacokinetics. Tumor shrinkage is seen in some patients, and an incremental improvement in survival occurs in the group. However, due to their large size and consequent slow diffusion, antibody penetration deep into tumors may be inhomogeneous. Even if only a few cells, deep in tumors, escape therapy, they can regrow and lead to clinical relapse, limiting the significant potential of monoclonal antibody therapy. This leads to questions about optimal dosing for monoclonal antibodies. Methods to determine monoclonal antibody dose include maximum-tolerated dose studies, pharmacokinetically and pharmacodynamically guided dosing, randomized dose-ranging studies, imaging of antibody biodistribution and competitive-binding studies. Limitations of these methods, and future directions to possibly overcome these limitations will be discussed.