We have recently developed a retroviral vector that contains a splice acceptor from the human EF1-a gene and drives a significantly higher level of gene expression than other well known murine leukemia virus-based vectors. However, one downside of this vector is that viral titer significantly varies depending on the packaging lines used. Results from Northern blot analysis indicated that in certain cell lines the genomic transcript containing the packaging signal sequence was too efficiently spliced to the subgenomic RNA, resulting in low levels of genomic RNA and thus leading to a low viral titer. We tested the possibility of overcoming this problem by introducing mutations around the splice acceptor sequence in such a way that a delicate balance was maintained between the splicing efficiency (which determines the level of gene expression) and the amount of genomic transcript (which influences viral titer). After mutational analysis, one such mutant was found to meet this requirement. The newly constructed vector containing the engineered splice acceptor could indeed drive higher levels of expression in many therapeutic genes than other control vectors, without significantly compromising viral titer.