Site-specific genetic engineering is a valuable tool for pharmaceutical research and development of biomedical models. Despite engineered nucleases allow targeted gene edition in a rather simple fashion; few reports are available so far on specific gene knock-in (KI) combined with engineered nucleases in domestic species. Here, we evaluated the possibility of inducing specific KI of cDNAs coding for proteins of pharmaceutical interest under the control of milk native promoter sequences, taking advantage of the TALEN system, both in ovine somatic cells and in zygotes. We designed 2 TALENs, targeting exons 1 and 5 of ovine β-lactoglobulin gene (BLG), respectively, and a homologous recombination vector (pHR), carrying recombinant human factor IX (rhFIX) flanked by homology arms contiguous to the TALEN target sites. In an initial set of experiments, 5 × 105 to 1 × 106 ovine fibroblasts were transfected with 1 μg of each TALEN mRNA, with or without 50 ng μL−1 pHR. The feasibility of inducing knock-out (KO) was confirmed by Cel1 assay. The deletion of the genomic region between TALEN target sites and the occurrence of HR in cell lysates were assessed by PCR. Also, 576 individual colonies were picked up and analyzed by PCR. The deletion of the region between TALEN target sites was achieved with 7.8% efficiency (45/576). The incidence of HR in cells was 0.5% (3/576), as detected by PCR. In order to evaluate the system in zygotes, laparoscopic AI was performed on synchronized and superovulated ewes. Zygotes were recovered 16 h after AI and cytoplasmically injected with (1) 5 ng μL−1 TALEN mix (2.5 ng μL−1 oaBLG T1.1 + 2.5 ng μL−1 oaBLG T5.1) (5TM); (2) 5 ng μL−1 TALEN mix + 25 ng μL−1 pHR-hFIX plasmid (5TM+25pRH); or (3) 15 ng μL−1 TALEN mix (7.5 ng μL−1 oaBLG T1.1 + 7.5 ng μL−1 oaBLG T5.1) + 50 ng μL−1 pHR-hFIX (15TM+50pRH). A non-injected control (NIC) was also included. Embryo analysis was conducted on whole-genome amplified DNA from blastocysts, followed by PCR and sequencing. Non-parametric Fisher test was applied to detect significant differences among treatments. Although blastocyst rates for NIC and 5TM did not statistically differ, 5TM+25pRH and 15TM+50pRH groups resulted in lower blastocysts rates than the NIC [P < 0.05; 13/17 (76%), 6/15 (40%), 4/15 (26%) and 2/14 (14%) for NIC, 5TM, 5TM+25pRH and 15TM+50pRH respectively]. It was possible to detect the PCR product compatible with deletion of the entire region among TALEN target sites in 6/6 blastocysts (100%) from the group 5TM, 3/4 blastocysts (75%) from the group 5TM+25pRH and 2/2 (100%) blastocysts from the group 15TM+50pRH. HR was detected in 1/2 (50%) blastocysts injected with 15TM+50pRH and in 1/4 (25%) blastocysts injected with 5TM+25pRH, by PCR and sequencing of the PCR products. Our results indicate that TALEN combined with homologous recombination constitutes a powerful platform for the production of proteins of pharmaceutical interest under native regulatory sequences in the milk of genetically modified animals.