This paper deals with the effect of bond thickness, embedment length and type of epoxy adhesive formulation on the adhesion of steel anchors to concrete. The test results indicate that the shear strength of the epoxy adhesive formulation prepared from a diglycidylether of bisphenol A (DGEBA) and an ethyleneamine curing agent is independent of bond thickness from approximately 1-4 mm. However, the same is not found to be true for the DGEBA epoxy resin cured using an alkylenediamine curing agent where considerably more complex behaviour is observed. The adhesive shear strength of this system is shown to pass through a maximum value at 2 mm bond thickness. At thickness beyond 2 mm, the strength undergoes a reduction until a value is reached which remains essentially constant with increased thickness. The variation in concrete strength is regarded as a main factor responsible for this complexity. The addition of filler to epoxy adhesive formulations results in a reduction in the adhesive shear strength, the magnitude of the decline being largely dependent upon the quantity of filler used in the mixture. The quadratic equations reviewed in this study provide a basis for assessing trends in the mechanical behaviour of filled epoxy systems. The shear strength of an epoxy adhesive anchor can be determined by an elastic behavioural model. The curves for epoxy adhesives indicate that the elastic solution does fit the experimental data very well. Increasing embedment of an anchor beyond a certain point does not increase the ultimate tension capacity of the anchor. (C) 2001 Elsevier Science Ltd. All rights reserved.