It has become apparent that the clinical success anticipated in the field of gene therapy has been limited by progress in several of the fundamental areas of genetics, molecular and cellular biology relevant to its application. Whilst a great deal of effort has been made in the evaluation of transgenes, it is only more recently with the advance of vector systems that attention has begun to be focused upon the means and control of transgene expression. Until recently, the majority of constructs have employed ubiquitous viral promoters to drive expression from simple gene expression cassettes using viral promoters and lacking introns, 3' untranslated regions (UTRs), locus control regions (LCR's), matrix attachment regions (MAR's) and other such genetic components. It has consequently emerged that these elements may have a key role in determining the levels and longevity of gene expression attainable in vivo, irrespective of the vector system utilised. The majority of gene therapy applications would also benefit from the specific optimisation of 'tailor-made' expression cassettes to optimise their therapeutic efficacy. In conjunction with modification of vector tropism and strategies to limit their immunogenicity, this should create vectors suitable for the clinical application of gene therapy. This review aims to highlight some of the principle considerations of gene expression in vivo, and the means by which it may most effectively be achieved, whether this is via the minimal modification of an existing eukaryotic promoter or by the more extensive design of a novel promoter and associated elements.