Purpose: To investigate the influence of composite resin consistency and placement technique on proximal contact tightness of Class II composite resin restorations. Materials and Methods: A manikin model (KaVo Dental) was used with an artificial first molar in which a standardized MO preparation was ground. This preparation was duplicated 360 times. Cavities were restored using Clearfil Photo Bond (Kuraray) combined with one of three composite resins of different consistencies: a low-viscosity (X-Flow, Dentsply), a medium-viscosity (Clearfil AP-X, Kuraray) and a high-viscosity composite (Tetric Ceram HB, Ivoclar Vivadent). Each composite was combined with 6 different matrix systems and separation techniques (n = 20). Groups 1 and 2: pre-contoured metal circumferential matrix (KerrHawe 1101-c) in a Tofflemire retainer combined either with hand instrument (OptraContact, Ivoclar Vivadent) or separation ring (Composi-Tight Gold, Garrison Dental Solutions). Group 3: pre-contoured metal sectional matrix (Lite-Flex, Danville Materials) with separation ring. Groups 4 and 5: precontoured metal circumferential dead-soft matrix (Adapt SuperCap, KerrHawe) with or without separation ring. Group 6: flat metal circumferential matrix (OptraMatrix, Ivoclar Vivadent) in a Tofflemire-retainer with hand instrument (OptraContact). Proximal contact tightness was measured using the Tooth Pressure Meter (University of Technology, Delft). To determine the effect of experimental variables on the proximal contact tightness, a multiple linear regression model was constructed. Results: Measurements in group 6 were not possible; therefore, this group was excluded. The use of medium- or high-viscosity instead of a low-viscosity composite resin resulted in statistically significantly tighter proximal contacts (p 0.01). The use of a separation ring resulted in a large, statistically significant increase (p 0.001) in contact tightness, while the use of a hand instrument resulted in a small, statistically significant increase of contact tightness (p = 0.017). No statistically significant differences were found when a dead-soft matrix or a sectional matrix was used instead of a Tofflemire (p = 0.159, p = 0.261, resp.). Conclusion: Use of a separation ring when restoring a Class II composite resin restoration has a greater influence on the obtained proximal contact tightness compared to the influence of the consistency of the composite resin. Keywords: proximal contact, Class II composite resin restoration, matrix system, separation ring