Studies aimed at elucidating mechanisms responsible for the expression of drug resistance, which in the past have centred largely on the use of experimental animal model tumour systems, have more recently employed human tumour material. The advantages and limitations of the two main approaches adopted for in vitro studies, which have involved either direct use of biopsy specimens or work with human tumours established as continuous cell lines, are reviewed here. Reference is made to the need for strict characterization of the tumour cell populations under investigation, and their potential instability and heterogeneity, particularly in terms of drug responses, are discussed. Studies using human tumour cell lines have been more extensive than those employing biopsy material directly. Examples of investigations analysing drug resistance using lines derived from gliomas and from carcinomas of the lung, ovary or colon are reviewed. The question of whether these human tumour model systems accurately reflect the types of resistance encountered clinically is considered. Finally, several potential areas for further study are proposed centering on current investigations that address the issues of whether different mechanisms of resistance might be involved depending not only on the order of resistance expressed but also on the way in which such resistance was developed or induced.