Shiga toxigenic Escherichia coli (STEC) are important foodborne pathogens causing gastrointestinal disease worldwide. Understanding how STEC attach to food-related surfaces may provide insight into their potential persistence and contamination within food environments. Bacterial attachment is influenced by a number of bacterial surface properties, including physicochemical characteristics, surface structures, protein expression and modes of growth. The aim of this study was to characterise the surface properties of 17 STEC and 3 non-STEC strains representing a range of serotypes and to determine the role of these surface properties in the initial attachment of STEC to beef muscle and adipose tissue and stainless steel. In addition, investigations into whether initial attachment of STEC to stainless steel could predict biofilm formation were also performed. The physicochemical properties of the STEC strains, namely cellular surface charge, hydrophobicity and electron donor/acceptor potential were determined for all strains grown in planktonic and sessile culture. Cellular surface charge was determined using zeta potential measurements. Hydrophobicity of the isolates was determined using three hydrophobicity determination methods, while a microbial adhesion to solvents method was used to determine the electron donor/acceptor characteristics. No differences of surface charge measurements were found between planktonic and sessile grown cultures. Isolates belonging to serogroup O157 and serotypes O26:H11 and O111:H- were significantly (p0.05) different between STEC isolates for all assays. For stainless steel, the attachment of STEC was determined using epifluorescence microscopy. Five STEC strains attached in significantly greater (p0.05) in attachment when grown in planktonic or sessile culture. The change of interfacial free energy between the bacterial strains and stainless steel was calculated and the influence of free energy in attachment was determined. Although a significant difference (p<0.05) in free energy values was found between STEC strains, no correlation was found between free energy values and bacterial counts on stainless steel. In addition, no correlation was found between bacterial hydrophobicity and surface charge values or production of surface structures (type I fimbriae or flagella) with viable counts and Sr values for muscle and adipose assays and for counts obtained for the stainless steel assays. In order to determine whether initial attachment can predict biofilm formation, two biofilm determination methods (epifluorescence microscopy method on stainless steel and a microtitre plate method) were performed and evaluated using seven STEC and three non- STEC (in diluted NB, 25°C, 24 and 48 h), grown in planktonic and sessile culture. The influence of surface structure expression (flagella, type I fimbriae and curli fimbriae) and the possession of antigen 43 (encoded by agn43) on biofilm formation were also investigated. No correlation was found between counts for epifluorescence microscopy on stainless steel and the absorbance values obtained with the microtitre plate method for planktonic and sessile grown cultures. Different abilities of individual STEC strains to attach to stainless steel and microtitre plates were found, with some strains attaching better to each surface following growth in either planktonic or sessile culture. All O157 STEC strains had low biofilm counts on stainless steel for planktonic and sessile grown cultures, however one STEC O157:Hstrain (EC516) had significantly greater (p<0.05) biofilm production on microtitre plates compared to the other O157 STEC strains. EC516 also had greater biofilm production on microtitre plates following growth in sessile culture in comparison to planktonic culture. Strains expressing curli fimbriae were found to produce significantly greater (p<0.05) biofilms on microtitre plates compared to the non-curli expressing strains. No relationship was found between the production of type I fimbriae, motility, agn43 and bacterial physicochemical properties and biofilm formation on stainless steel or microtitre plates. The outer membrane protein (OMP) profiles of four E. coli O157 and one non-STEC strain demonstrating different abilities to initially attach to stainless steel following growth in planktonic and sessile culture were obtained using 2D electrophoresis. Qualitative and quantitative differences in the total number of OMPs expressed between planktonic and sessile cultures were found for all E. coli isolates tested. Two of the five E. coli O157 isolates were found to express a greater number of proteins in sessile culture (80-111 proteins) compared to planktonic culture (53-54 proteins). The number of proteins expressed only in sessile culture was also greater (15-44 proteins) than the number of proteins expressed in planktonic culture only (3-7 proteins). Several proteins were also found to be up-regulated in either planktonic (3-7 proteins) or sessile culture (3-8 proteins), in comparison to the other mode of growth tested. Differentially expressed proteins spots were identified using matrixassisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF-MS). A flagella protein (FlgE) was found to be expressed in planktonic culture only, while ironregulating proteins (FhuA, FepA and YbiL) were expressed in sessile culture only. Other OMPs (MipA and OmpX) had lower expression in sessile culture than in planktonic culture The results of this study demonstrate that the surface properties of STEC strains (e.g. physicochemical characteristics, surface structure and OMP expression) can differ between strains grown in planktonic and sessile culture. STEC were also found to behave differently with respect to attachment to meat tissues and stainless steel, with some strains demonstrating a greater potential to strongly attach to a surface following growth in planktonic or sessile culture. Initial attachment of STEC to stainless steel does not appear to influence biofilm formation and differences between biofilm determination methods tested suggests that biofilm production on one surface may not be appropriate to represent other surfaces. A lack of correlation between the surface properties characterised and the attachment of STEC to surfaces tested in this study suggests that STEC may use a combination of known and unknown attachment mechanisms which in turn are influenced by the bacterial cell, the attachment surface and surrounding medium. Further investigations are needed to continue characterising bacterial and substratum factors involved in bacterial attachment and biofilm processes.