Poor repeatability of peak areas is a problem frequently encountered in peptide analysis with nanoLiquid Chromatography coupled on-line with Mass Spectrometry (nanoLC-MS). As a result, quantitative analysis will be seriously hampered unless the observed variability can be corrected in some way. Currently, labeling techniques or addition of internal standards are often applied for this purpose. However, these procedures are elaborate and error-prone and may render complex samples even more complex. Moreover, whenever poor repeatability results from variable recovery, not just quantification, but also sensitivity is affected. We have studied the parameters influencing the repeatability of chromatographic peak areas for a model set of proteolytic peptides (i.e., a cytochrome c tryptic digest) in nanoLC-MS analysis. It is demonstrated that repeatability issues are mainly due to poor recovery of peptides from the sample vial. Problems are largely resolved by addition of an organic modifier to the sample vial to improve solubility of the peptides, but care needs to be taken not to lose peptides due to reduced affinity for reversed-phase materials. Good results are obtained when applying dimethylsulfoxide (DMSO) for this purpose. When applying DMSO, repeatability increases, and the limit of detection (LOD) decreases. For the most hydrophobic peptides, a gain in LOD of at least an order of magnitude is obtained. In an aqueous sample containing 0.1% formic acid (FA), it is possible to detect 100-200 fmol of peptide, whereas +/-10 fmol can be detected in a sample containing 5% FA and 25% DMSO (10 microL injections).