Using a selection process designed to reflect clinically relevant conditions, a panel of taxane-selected variants were developed to study further the mechanisms of resistance in lung cancer. Unlike continuous or pulse exposure to high concentrations of chemotherapeutic drugs which yield high resistance and often cross resistance, most variants developed here displayed low level resistance to the selecting drug with slight cross-resistance. Pulsing with taxol resulted in more highly resistant clones (up to 51.4-fold). Analysis of taxol and taxotere in the four major lung cancer cell types showed the taxanes to be more effective against NSCLC (with the exception of SKMES-taxane selected variants) than against the SCLC. Comparison of taxol and taxotere shows that taxol induces higher levels of resistance than taxotere. Further, in taxotere-selected cell lines, the cells are more resistant to taxol than taxotere, suggesting that taxotere may be a superior taxane from a clinical view. Taxol treatment resulted in increased cross-resistance to 5-FU in all classes of lung cancer except DMS-53. The high levels of Pgp in the DMS-53 and selected variant suggests this mechanism is not related to Pgp expression. Analysis of the Pgp and MRP-1 status by combination inhibitory assays and Western blotting showed no consistent relationship between expression of the membrane pumps Pgp or MRP-1 and resistance. However, where high level resistance was seen, the parent cell line expressed Pgp or MRP-1 and was accompanied by increased levels in the variants. Overall we found that the clinically relevant models used here are useful for investigating mechanisms of taxane resistance.