Protein-spanning peptide pools have proven valuable as a screening tool for detecting T-lymphocyte responses against a wide range of proteins. We have used this approach in our search for T cells reactive to the onconeural protein HuD. We found positive responses in only 3 of 127 individuals; however, these were highly unusual in that the same class I HLA alleles and peptides were involved. These T-cell responses were not confirmed when peptides re-synthesized by the same manufacturer with similar and with higher purity levels were used. Our observations indicated that these T-cell responses were not directed against the designed HuD peptides. Here, we report on (i) comparisons of the peptide batches analyzed by matrix-assisted laser desorption/ionization Fourier transform mass spectrometry (MALDI-FTMS) that did--and did not--elicit T-cell responses and (ii) a detailed analysis of the various by-products of peptides, irrespective of T-cell assay outcome. We found numerous differences between the peptide batches, such as omissions of amino acids in the primary structure of the peptides. Furthermore, some batches revealed strong interactions with calcium ions or contained sulfated peptides. Our data reveal that different batches from the same peptide may contain artefacts that influence the outcome of HLA-restricted T-cell response assays.