Immunotherapy based on whole cancer cell vaccines is regarded as a promising avenue for cancer treatment. However, limited efficacy in the first human clinical trials calls for more optimized whole cancer cell vaccines and better patient selection. It is suggested that whole cancer cell vaccines consist preferably of immunogenically killed autologous cancer stem cells associated with dendritic cells. Adjuvants should stimulate both immune effector cells and memory cells, which could be achieved through their correct dosage and timing of administration. There are indications that whole cancer cell vaccination is less effective in patients who are immunocompromised, who have specific genetic defects in their immune or cancer cells, as well as in patients in an advanced cancer stage. However, such patients form the bulk of enrolled patients in clinical trials, prohibiting an objective evaluation of the true potential of whole cancer cell immunotherapy. Each key point will be discussed.