When limbal stem cell deficiency (LSCD) is partial, the usual treatment is covering the corneal surface with amniotic membrane (AM), allowing the growth of the remaining limbal stem cells (LSCs). In cases of complete LSCD, the most common treatment is cultured limbal epithelial transplantation (CLET), with the possibility of rejection. Studies proved that mesenchymal stem cell (MSC) transplantation is as safe and effective as CLET. Recent studies demonstrate a successful differentiation of adipose-derived adult mesenchymal stem cells (ADSCs) to LSCs. Combining AM transplantation with LSCs results in improved treatment efficacy. A disadvantage of using AM is the variability among donors and the risk of rejection. We propose the use of 3D-printed collagen as a scaffold colonized with LSCs derived from ADSCs for the treatment of LSCD in a rat animal model. The 3D-printed collagen scaffolds showed good transparency. In vitro ADSC differentiation to LSCs showed a change in cell morphology, being more pronounced and faster on 3D-printed collagen. 3D-printed collagen was the most efficient differentiation substrate, with the highest percentage of LSC specific markers (p63α and BMI-1) and corneal epithelium (SSEA-4).  The differentiated LSCs in AM or 3D-printed collagen I scaffolds were transplanted into a rat model of LSCD, compared to the standard cell free AM. In all groups the epithelium closed the imposed wound. 3D-printed collagen scaffold integration was statistically superior to AM. Markers for different corneal structures (PAS, BMI-1, p63α, cytokeratins 12 and 13) showed that the generated epithelium was not epithelial but conjunctival, and that the contribution of ADSC-derived LSCs was not enough for re-epithelization of the cornea.