In this paper, we review recent work on multidrug resistance (MDR) in Amsterdam. We have generated mice homozygous for a disruption of one of their P-glycoprotein (Pgp) genes. The mutations do not interfere with viability or fertility, showing that these Pgps have no indispensable role in early development or metabolism. Mice homozygous for a disruption of their mdr2 gene, however, develop liver disease and this appears to be due to their complete inability to secrete phospholipids into bile. This suggests that the mdr2 Pgp (and, by inference, its human MDR3 homologue) is essential for translocating phospholipids through the hepatocyte canalicular membrane in which this Pgp is located. These and other results show the importance of the genetic approach for studying drug metabolism. MDR is not only caused by increased activity of Pgps. When the human non-small cell lung carcinoma cell line SW-1573 is selected in vitro for low level doxorubicin resistance, the resistant variants are nearly always multidrug resistant, but this is not due to increased Pgp activity. Only when resistance is pushed to higher levels does activation of the MDR1 Pgp gene occur. This suggests that clinically relevant levels of drug resistance in some cells may be caused predominantly by non-Pgp-mediated drug resistance mechanisms. The protein responsible for MDR in the SW-1573 cells has not yet been identified and experiments are in progress to find the gene encoding it.