Biomarkers in amniotic fluid (AF) include both non-modified and phosphorylated proteins and can be used in the diagnosis of pregnancy-associated pathologic conditions. In this work, an integrated LC–MS method for selective, sensitive and reproducible analysis of phosphorylation in proteins has been applied to AF. Online digestion of (phospho) proteins was coupled with the selective enrichment on a TiO2 trap, and separated by RPLC–MSn of both normal and phosphorylated produced peptides. First, an AF-pooled sample was analyzed and a general map of contained proteins and biomarkers was derived in a single run. Then, individual AF samples were analyzed with a downscaled platform with improved sensitivity. On purpose, a trypsin-based CIM minidisk was used for online digestion of AF. The obtained protein profile was highly consistent with the one obtained with traditional off-line digestions. Moreover, the use of a specific phospho-enrichment tool followed by LTQ-Orbitrap, enhanced the confidence in the determination of protein phosphorylation state and phosphorylation sites. The phosphorylation sites of IGFBP-1 and osteopontin present in the AF of two individual samples were monitored with a total of 24 and 17 phosphopeptides, respectively, encoding for 12 putative novel phosphorylation sites in addition to known sites.