Optimization of peptide stability is essential for the development of peptides as bona fide alternatives to approved monoclonal antibodies. This is clearly the case for the many peptides reported to antagonize proprotein convertase subtilisin-like/kexin type 9 (PCSK9), a clinically validated target for lowering cholesterol. However, the effects of optimization of stability on in vivo activity and particularly the effects of binding to albumin, an emerging drug design paradigm, have not been studied for such peptide leads. In this study, we optimized a PCSK9 inhibitory peptide by mutagenesis and then by conjugation to a short lipidated tag to design P9-alb fusion peptides that have strong affinity to human serum albumin. Although attachment of the tag reduced activity against PCSK9, which was more evident in surface plasmon resonance binding and enzyme-linked immunosorbent competition assays than in cellular assays of activity, activity remained in the nanomolar range (∼40 nM). P9-alb peptides were exceptionally stable in human serum and had half-lives exceeding 48 h, correlating with longer half-lives in mice (40.8 min) compared to the unconjugated peptide. Furthermore, the decrease in in vitro binding was not deleterious to in vivo function, showing that engendering albumin binding improved low-density lipoprotein receptor recovery and cholesterol-lowering activity. Indeed, the peptide P9-albN2 achieved similar functional endpoints as the approved anti-PCSK9 antibody evolocumab, albeit at higher doses. Our study illustrates that optimization of stability instead of binding affinity is an effective way to improve in vivo function.