Measurement of humoral tumor-specific immunity can predict what proteins are specific tumor antigens, be used to evaluate patient diagnosis or prognosis, and function as a method by which one can measure the effects of an immune intervention, such as a vaccine. Antibody assays can easily be adapted to high throughput formats; however, specific reagents needed for assay development often are not available. Developing methods to produce large quantities of purified recombinant tumor antigen proteins for indirect ELISA is both laborious and expensive. In addition, using proteins derived from E. coli might preclude the detection of certain antibody epitopes. We questioned whether a human tumor cell-based ELISA could be developed to assess antibody immunity to common tumor-associated antigens and whether such an ELISA could be optimized to the clinical standards needed for evaluation of large scale trials. Assays were based on the detection of HER-2/neu and p53 antibodies by capture ELISA, using human tumor cell lysate as a protein source. After optimization, the HER-2/neu and p53 ELISA intra-assay coefficients of variation (CV) of positive control sera were consistently 9% and 12%, respectively, at a 1:100 dilution. The HER-2/neu and p53 inter-assay CV of positive control sera over a 5-month time period were 20% and 15%, respectively. The sensitivity and specificity of the ELISAs were evaluated based on comparison to immunoblot. Analysis demonstrated the HER-2/neu ELISA had a specificity of 77% and sensitivity of 89%, and the p53 ELISA had a specificity of 100% and sensitivity of 93%. Cell-based ELISA can be developed to be Clinical Laboratory Improvement Act (CLIA)-compliant and the flexibility of the approach will allow adaptation of the assay to multiple tumor antigen systems.