Statement of problem: Deep subgingival defects at the abutment teeth for fixed dental prostheses (FDPs) are a common problem in prosthodontics. Surgical crown lengthening or orthodontic extrusion of the abutment teeth have been recommended treatment options. However, disadvantages such as extended treatment time, treatment costs, or surgical problems have to be considered. Proximal box elevation has been proposed as an alternative, but microleakage studies are lacking. Purpose: The purpose of this in vitro study was to investigate the influence of proximal box elevation (PBE) and conventional foundation restoration on the microleakage of 3 mol% yttria-stabilized tetragonal zirconia polycrystal (3Y-TZP) ceramic FDPs after thermomechanical loading (TML). Material and methods: Sixteen molars and 48 premolars were divided into 4 groups (n=8). Either fixed-to-fixed (premolar, molar) or cantilever FDPs (premolar, premolar) were fabricated. The pontic of both groups had a mesiodistal width of 7 mm. The proximal foundation restorations were placed using a self-etch adhesive system and a dual-polymerizing composite resin, and a standardized tooth preparation design was carried out. All FDPs were adhesively luted and subjected to TML (1.2 million cycles, 5/55 °C, 49 N). After TML, all specimens were placed in fuchsin and then sectioned in a sagittal direction. The sections were evaluated with respect to the penetration depth of the dye by light microscopy at 4 different interfaces: tooth and luting composite resin (TC), luting composite resin and restoration (CR), tooth and foundation restoration (TB), foundation restoration and luting composite resin (BC). As data were not normally distributed, nonparametric tests were used for statistical analysis (α=.05). Results: Statistically significant differences were observed between the PBE and the control group (P<.05). Microleakage at the TB interface was detectable in all specimens of the PBE group, while 84.4% of the specimens of the control group showed leakage at the TC and TB interfaces. Leakage was detected at the CR interface in 3.1% of all specimens. No leakage was observed at the interface BC. Additionally, statistically significant differences were found within the same study specimen between the control (conventional foundation restoration) and the study side (PBE) (P<.05). Conclusions: Irrespective of the pretreatment (experimental PBE or conventional foundation restoration), the adhesive bond of the foundation restoration and the luting composite resin to the dentin represented a limiting factor with regard to microleakage in fixed-to-fixed and cantilever FDPs. PBE was more frequently affected by microleakage than a conventional foundation restoration.