The therapeutic potential of antisense oligonucleotides will heavily depend on a balance of two factors: pharmacologic effectiveness and cost of production. Pharmacologic optimization will be achieved to a limited degree in in vitro systems, but substantial progress can only be made in the context of appropriate in vivo models. The quantities of synthetic oligonucleotides required for modest in vivo testing are several thousandfold greater than can be produced by conventional DNA synthesis technology and 10(5)-10(7)-fold greater for preclinical and clinical evaluation. Cost-effective synthesis and purification cannot be achieved by extrapolating current technologies to scales commensurate with these quantities. Recent interest in anti-HIV (anti-rev) phosphorothioate analogues of DNA (approximately 28-mers) has prompted us to develop scale-up methodology for routinely producing gram amounts of such analogues. These results and considerations given to producing clinical and commercial quantities were discussed.