We reviewed toxicological studies, both experimental and epidemiological, that appeared in international literature in the period 1990-1997 and included both leaded and unleaded gasolines as well as their components and additives. The aim of this overview was to select, arrange, and present references of scientific papers published during the period under consideration and to summarize the data in order to give a comprehensive picture of the results of toxicological studies performed in laboratory animals (including carcinogenic, teratogenic, or embryotoxic activity), mutagenicity and genotoxic aspects in mammalian and bacterial systems, and epidemiological results obtained in humans in relation to gasoline exposure. This paper draws attention to the inherent difficulties in assessing with precision any potential adverse effects on health, that is, the risk of possible damage to man and his environment from gasoline. The difficulty of risk assessment still exists despite the fact that the studies examined are definitely more technically valid than those of earlier years. The uncertainty in overall risk determination from gasoline exposure also derives from the conflicting results of different studies, from the lack of a correct scientific approach in some studies, from the variable characteristics of the different gasoline mixtures, and from the difficulties of correctly handling potentially confounding variables related to lifestyle (e.g., cigarette smoking, drug use) or to preexisting pathological conditions. In this respect, this paper highlights the need for accurately assessing the conclusive explanations reported in scientific papers so as to avoid the spread of inaccurate or misleading information on gasoline toxicity in nonscientific papers and in mass-media messages.