SummaryTo form distatit metastases. tumour cells must stabilize adhesive interactions that prevent detachment at secondary sites. Primary receptor-ligand interactions alone may not maintain prolonged adhesive contacts without secondary events that lead to adhesion stabilization. Computerized imaging methods enable us to examine various substrates for: (i) the wall shear adhesion threshold (WSAT). a measure ofthe dynamic adhesive potential of tumour cells; (ii) the number of tumour cells that adhered; and (iii) the adhesion stabilization lag time (ASLT) or length of time required for tumour cells to stabilize adhesive contacts capable of withstanding high wall shear force (up to 100 dynes/cm-). The relative WSAT ratios found were: wheat germ agglutinin (WGA) > laminin > fibroncctin > vitronectin > collagen I > collagen IV > von Willebrand factor (vWF) (the greater the shear rate the higher the adhesive potential). The relative stabilization ratios found were as follows: laminin < fibronectin < vitronectin < collagen IV < collagen I < vWF < WGA (shorter times correlate with greater stabilization potential). Stabilization data using fibronectin as a substrate correlated the best with metastatic potential. Using three melanoma lines of ditterent metastatic potential semiquantitative reverse transcriptase-polymerase chain reaction (PCR) showed a two-to four-fold increase in al, a3, a4. a5. a6, and ICAM-1 in the highly metastatic 70W celts compared to the MeWo and non-metastatic 3S? melanoma cells. There were no differences in av. p I and p3 levels among the three melanoma lines, and PCR products for allb. a2, CD36. or ICAM-2 were not detected. The 70W cells also had higher levels of ax and p2 (CD11/CD18 and pl50 leukocyte antigen) than either the MeWo or 3S5 cells. The data indicate that melanoma cells exhibit differences in the adhesion properties under fluid shear and differences in the expression of adhesion components that correlate with their metastatic potential.