The utility of CEA and CA19-9 as colorectal carcinoma (CRC) markers is limited and development of additional reliable markers is under investigation. We previously showed that galectin-1 is overexpressed in CRC tissues. If such a protein leaks into the peripheral circulation, it might constitute a tumor marker candidate. Here, we test the hypothesis that the levels of circulating galectins could reflect the presence of CRC and/or its progression state. We constructed sandwich ELISAs for galectin-1/-2/-3/-4/-7 and determined their plasma concentrations in 105 CRC patients and 100 healthy volunteers (control). Matched pair samples of 56 patients pre- and post-surgery were also subjected to ELISA analysis. Circulating levels of galectin-1/-3/-4 in CRC patients were significantly higher compared to those in controls. Galectin-1 and galectin-4 levels significantly decreased after surgery (P<0.01), and the level of galectin-4 in most patients fell below the cut-off value. The levels of circulating galectin-4 significantly increased as the tumor stage progressed (P<0.001), whereas those for galectin-1 were relatively high from an early stage. Combined use of galectin-4 with CEA and/or CA19-9 markedly increased the proportion of CRC patients who were positive for tumor markers (from 33.3 to 59.0% for CEA and from 17.1 to 51.4% for CA19-9). Our data show that galectin-4 may be a tumor marker for use in patient follow-up, while galectin-1 could be used for tumor screening. In particular, galectin-4 can be useful as a complementary marker when combined with CEA/CA19-9 to improve CRC follow-up.