Urinary trypsin inhibitor (UTI) inhibits efficiently tumor cell invasion and the formation of metastasis. The anti-metastatic effect is dependent on the COOH-terminal domain II of UTI [UTI-(78-136)-peptide]. To develop a molecule that binds with high affinity to the urokinase (uPA) receptor (uPAR) on tumor cell surfaces, a bifunctional hybrid molecule [uPA-(1-134)-UTI-(78-136)] consisting of the uPAR-binding NH2-terminal fragment [UTI-(78-136)-peptide] of uPA at the NH2-terminus of UTI-(78-136)-peptide was produced in Escherichia coli by genetic engineering. The purified hybrid protein inhibited trypsin and plasmin 2-3-fold less effectively than UTI-(78-136)-peptide and was found to bind to human tumor cells via uPAR, which was confirmed by cell binding and competition experiments. Using a modified Boyden chamber and an artificial basement membrane, Matrigel, it was found that the hybrid protein is very effective at inhibiting invasion by uPAR-expressing human tumor cells. Sensitivities of tumor cells towards the anti-invasive effect of uPA-(1-134)-UTI-(78-136) correlated with the density of uPAR on human tumor cells. Furthermore, in the spontaneous metastasis model, the hybrid protein inhibited the formation of lung and/or lymphatic metastasis by human ovarian carcinoma and choriocarcinoma cells. The hybrid protein was much more effective than uPA-(1-134)-peptide, UTI-(78-136)-peptide, or UTI. We conclude that this approach extends the possibility of applying recombinant protein for therapeutic use in inhibition of human tumor cell metastasis.