This study investigated the tightness of the proximal contact when placing posterior resin composite restorations with circumferential and sectional matrix systems in an in vitro model using a special measuring device (Tooth Pressure Meter). A manikin model was used with an artificial first molar in which an MO-preparation was ground, simulating the clinical situation of an amalgam replacement. This preparation was duplicated, resulting in 160 identically prepared teeth. These teeth were divided into 8 groups (n=20). In 2 groups, circumferential matrix bands (flat or contoured) in a Tofflemire retainer were applied. In the remaining 6 groups, 3 different separation rings were combined with 2 types of sectional matrix bands. All the cavities were restored using Clearfil Photo Bond and Clearfil AP-X. The tightness of the proximal contact was measured using the Tooth Pressure Meter. Data were statistically analyzed using SPSS 12. ANOVA was used to find differences in proximal contact tightness between the groups. Tukey tests were used to find differences between the homogeneous subgroups. The use of sectional matrices combined with separation rings resulted in tighter proximal contacts compared to when circumferential systems were used (p<0.001). The use of these devices is therefore recommended when posterior resin composite restorations are placed.