Recent political developments and disclosures of serious adverse events in human gene therapy (HGT) with the death of 18-year old Jesse Gelsinger in the USA have shown that the clinical application of HGT raises some severe ethical issues. These have either been neglected or not yet been discussed to a satisfactory extent. In this paper, we will address this deficiency and develop strategies for a safer application of HGT. Such a study must first look closely at the science of HGT itself. We will evaluate the latest preclinical research, especially data on the viruses that are used as vectors and on modes of administration of vectors. We will put forward new arguments concerning the toxicity assessment of so-called 'gene drugs', the tissue and cell type specificity of the vectors, and the duration and on-set of gene expression. Secondly, we will look at procedural aspects of applied research ethics on the way to clinical application of HGT. There, informed consent (IC) and the patient-researcher relationship are of utmost concern. Furthermore, we will explore the problem of expertise in risk assessment and will show how current regulations foster conflicts on interests that create dilemma situations even for those researchers who act in the best interest of the patients. We will conclude the article with a set of questions for ethicists who have to decide about the quality of HGT protocols. This may contribute to the safety of patients participating in HGT trials and to achieving the aim of efficient application of HGT.