Collagen prostheses used to repair abdominal wall defects, depending on their pretreatment (noncross-linked vs. cross-linked), besides repair may also achieve tissue regeneration. We assessed the host tissue incorporation of different bioprostheses using a new tool that combines immunofluorescence confocal microscopy with differential interference contrast images, making it possible to distinguish newly formed collagen. Partial hernial defects in the abdominal wall of rabbits were repaired using cross-linked/noncross-linked bioprostheses. Expanded polytetrafluoroethylene (ePTFE) was used as control. After 14/30/90/180 days of implant, specimens were taken for microscopy, immunohistochemistry, and quantitative-reverse transcription-polymerase chain reaction to determine host tissue ingrowth and collagen I/III protein and 1a1/3a1 gene expression. Shrinkage and stress resistance were also examined. At 14 days, cross-linked prostheses had suffered significantly less shrinkage than ePTFE or noncross-linked prostheses. Significantly higher shrinkage was recorded for ePTFE in the longer term. Microscopy revealed encapsulation of ePTFE by neoformed tissue, while the bioprostheses became gradually infiltrated by host tissue. Noncross-linked prosthesis showed better tissue ingrowth, more intense inflammatory reaction and more rapid degradation than the cross-linked prostheses. At 14 days, cross-linked prostheses induced up-regulated collagen 1a1 and 3a1 gene expression, while noncross-linked only showed increased collagen III protein expression at 90 days postimplant. At 6 months, the tensile strengths of cross-linked prostheses were significantly greater compared with ePTFE. Our findings demonstrate that despite the cross-linked collagen prostheses promoting less tissue ingrowth than the noncross-linked meshes, they became gradually replaced by good quality host tissue and were less rapidly degraded, leading to improved stress resistance in the long term.