OBJECTIVE: A PE analysis was performed to support the reimbursement request of erlotinib in 2nd/3rd-line treatment of NSCLC in the Netherlands (NL). METHODS: Erlotinib and BSC efficacy data (based on the erlotinib registration study, BR.21) were used for this analysis. Chart reviews (n=96) were conducted to obtain insight into health care utilisation (HCU) of stage IIIB/IV relapsed NSCLC. Charts from patients treated with docetaxel (n=24) and BSC (n=72) in 4 general and 1 academic hospital were used. The PE analysis was performed from the societal perspective and both outcomes and efficacy results were discounted at 4%. Official price lists (2004) were used and the price of erlotinib was set at €2184/150mg/30 tablets. PE outcomes extrapolated to 3 years were evaluated using a Markov health-state model, adapted for NL. Outcomes and model assumptions were approved by an expert panel of 10 Dutch clinicians. RESULTS: The average treatment costs per patient in NL were €24,939 for docetaxel, €23,436 for erlotinib, and €15,450 for BSC. Life-years gained (LYG) were 0.84 years for docetaxel and erlotinib and 0.62 years for BSC, as per the BR.21 registration trial intent-to-treat population. The incremental cost-effectiveness ratio (ICER) for erlotinib vs BSC was €37,059/LYG (CI €12,621–€72,960) based on 4.3 month treatment duration. Erlotinib dominated docetaxel in all scenarios except when an unrealistically low docetaxel dose (110mg/cycle) was assumed. ICERs were sensitive to variations in length/frequency of hospitalizations and number of outpatient visits, illustrating the economic impact of erlotinib's generally mild adverse event profile. Erlotinib was cost-effective vs BSC in 80% of cases using a willingness-to-pay (WTP) threshold of €50,000/LYG. CONCLUSIONS: Treatment with erlotinib dominates docetaxel and is cost-effective vs BSC in NL. Based on the clinical efficacy and cost-effectiveness, erlotinib has received unrestricted reimbursement for relapsed NSCLC in NL without requirements for patient stratification.