Plants with transgenic plastid (chloroplast) genomes represent a promising production platform in molecular farming, mainly because of the plastids’ potential to accumulate foreign proteins to very high levels and the increased biosafety conferred by the maternal mode of plastid inheritance. Although some transgenes can be expressed to extraordinarily high levels, the expression of others has been unsuccessful. Lack of detectable transgene expression is usually attributable to either RNA instability or protein instability. Here, we have investigated the possibilities to improve the production of a pharmaceutical protein that is difficult to express in chloroplasts: the HIV-1 fusion inhibitor cyanovirin-N (CV-N). Testing various N-terminal and C-terminal fusions to peptide sequences from two proteins known to accumulate to high levels in transgenic plastids (GFP and the protein antibiotic PlyGBS), we show that both low mRNA stability and low protein stability contribute to the lack of detectable CV-N expression in chloroplasts. Both problems can be alleviated by N-terminal fusions to the CV-N coding region, thus highlighting a suitable strategy for optimization of plastid transgene expression.