This study sought to determine whether protein and/or peptide profiles from serum were able to distinguish patients suffering from depression from healthy control subjects and thereby act as biomarker candidates with potential diagnostic value. Serum samples were collected from patients (n = 39) and controls (n = 30). A C8 magnetic bead protocol was used to prepare serum proteins, while a microextraction C18 packed tip was used to isolate serum peptides. Both protein and peptide profiles were recorded by MALDI-MS and the data were exported for further analysis. No protein signals differentiated patients from controls and principle component analysis of the entire peptide profile did not allow for distinct clustering of the two groups. Further analysis of individual peptides however identified three peptide signals whose intensities were significantly different between patients and control subjects. The efficacy of these potential biomarker candidates to identify the patients was therefore studied using a receiver operating characteristic (ROC) curve and the combined use of all three candidates together offered the most specific and sensitive identification of true positive cases of depression.