Small-cell lung cancer (SCLC) is an aggressive form for cancer with good treatment options, but often detected in a late stage. Pro-gastrin releasing peptide (ProGRP) is one of the biomarkers for SCLC and has traditionally been detected with immunoaffinity methods. The use of a bottom-up proteomics approach and LC-MS/MS can be an alternative to the immunological techniques. As ProGRP is a low abundance biomarker, a sensitive and selective method for sample preparation, both clean-up and target enrichment, is needed. Magnetic molecularly imprinted polymers (MMIPs) are magnetic beads generated to recognize and bind one specific target molecule, in this case the tryptic signature peptide of ProGRP, NLLGLIEAK. The use of MMIPs for a selective and sensitive sample preparation method before LC-MS/MS may therefore be an option. To evaluate the use of MMIPs as a tool in sample preparation of ProGRP two different MMIPs, MMIP A and MMIP B, were investigated. Both were fully magnetized beads with divinylbenzene (DVB) as the cross-linker. Functional monomer of MMIP A was N-(2-aminoethyl) methacrylamide hydrochloride (EAMA), while MMIP B had EAMA and a urea monomer as functional monomers. For MMIP B, in-sufficient elution recoveries (32 % maximum, RSD 20 %) were obtained. In addition, MMIP B showed poor inter-day reproducibility of amount target peptide bound (variations from 50-100 %). Optimization of the method for MMIP A was successfully done by evaluation of the following parameters: composition of test solution for binding, binding time, organic content in wash solution and composition of elution solution. The optimized method consisted of the following steps: binding time of 5 minutes in aqueous environment, wash with pure water (5 min) and 1 hour elution with 80:15:5 acetonitrile:water:formic acid (ACN:H2O:FA). The MMIP was also characterized with binding isotherms of the MMIP and the corresponding magnetic non-imprinted polymer (MNIP), and a successful imprinting and a clear difference between MMIP and MNIP with an imprinting factor of 6.12 was determined. Testing of the optimized method with a ProGRP digest was performed (> 90 % bound and good elution recoveries (87 %, RSD < 10 %) of target peptide). Furthermore, the method demonstrated good selectivity with small amounts (< 8 %) of other ProGRP peptides in the eluates. The optimized method was also tested in serum samples spiked with signature peptide with promising results, but further verification using ProGRP spiked serum is needed.