The use of osseointegrated titanium implants has become a widely accepted method to obtain retention for prosthetic reconstructions of partially dentate and edentulous patients. A range of c1inical long-term studies have been conducted to collect information on the numbers of surviving and/or successful implants for different implant systems. Overall implant success rates for several implant systems have thus repeatedly been reported to be as high as 95 % over 5 years and 90 % over 10 years ( Adell et al. 1981; Babbush et al. 1986; Albrektsson et al. 1988; Jemt et al. 1989; Adell et al. 1990; van Steenberghe et al. 1990; Henry et al. 1992; Mericske-Stern et al. 1994; Buser et al. 1997). A review of the above mentioned studies show that there are both common and system specific technical and biological complications and failures in implant prosthodontics. In the technical sector, a common observation has been that for almost all implant systems rigid connections between implants as well as between implants and teeth have turned out to be more favourable than non-rigid such connections. For the great majority of modern implant systems, rigid connections also seem to be associated with similar technical risk levels. The group of associated background factors inc1ude e.g. design features such as the extensions and dimensions of the restorations and patient specific factors such as bruxism (Aeschlimann et al. 1998; Brägger et al. 2001). Bruxing patients are at increased risk for complications in all restorative procedures. However, bruxism per se is not an absolute contraindication for dental implant therapy. Preventive measures such as the use of a habit appliance and protective nightguard should be considered. Certain specific material characteristics have also been noted. Especially over longer periods of time it has thus been reported that ceramic implants seem to be associated with high frequencies of severe mechanical complications such as fractures (Lill et al 1993). As has recently been pointed out by Brägger et al. (2001) in addition to common complications a range of system specific ones can also frequently be observed. Screw retained implant systems are thus associated with a higher frequency of loss of retention than cemented ones, although this may be balanced by possible lower frequencies for other types of technical failures. The cemented approach has the potential for being more passive than screw retention. Solid scientific and evidence-based data are missing within this field of dentistry (Nilner & Lundgren 1999). The question of whether a poorly fitting prosthesis can cause failure of osseointegrated dental implants has not been answered (Carr et al. 1996). Mollersten et al. (1997) experimentally evaluated how different joint designs influence the strength and the failure mode of dental implant systems. The strength tests focussed on the system as a whole (implant and connected abutments). Seven implant systems with different joint design were evaluated. Strength and failure mode varied significantly between the implant systems. Deep intemal joints, in contrast to shallow internal joints, favoured resistance to bending moments. A deep joint is one in which the screw takes little load and provides intimate contact with the implant walls to resist micromovement, resulting in a strong stable interface. Thus Mollersten et al. (1997) c1early demonstrated the strength advantage of an internal connection. They conc1uded that joint depth should be taken into consideration in construction of implant systems. Biological complications in the form of soft tissue reactions (periimplantitis) have been reported comparatively frequently in the field of implant prosthodontics, hut at present it is not certain if there exist any system specific risks (Wennström & Palmer 1999).