In past decades, studies focusing on new chemotherapeutic agents for patients with inoperable non-small-cell lung cancer have reported only modest gains in survival. These health gains are achieved at considerable cost, but economic evidence is lacking on superiority of one agent in terms of cost effectiveness. The objective of this systematic review was to assess fully published cost-effectiveness studies comparing the new agents docetaxel, paclitaxel, vinorelbine, gemcitabine and pemetrexed, and the targeted therapies erlotinib and gefitinib with one another. We performed systematic searches in the bibliographic databases PubMed, EMBASE and Health Economic Evaluations (HEED) [via the Cochrane Library] for fully published studies from the past 10 years. Studies were screened by two independent reviewers according to a priori inclusion criteria. The methodological quality of the included studies was evaluated by two independent reviewers using standardized assessment tools. A total of 222 potential studies were identified; 11 studies and six reviews were included. The methodological quality of the full economic evaluations was fairly good. Transparency in costs and resource use, details on statistical tests and sensitivity analysis were points for improvement. In first-line treatment, gemcitabine+cisplatin was cost effective compared with other platinum-based regimens (paclitaxel, docetaxel and vinorelbine). In one study, pemetrexed+cisplatin was cost effective compared with gemcitabine+cisplatin in patients with non-squamous-cell carcinoma. In second-line treatment, docetaxel was cost effective compared with best supportive care; erlotinib was cost effective compared with placebo; and docetaxel and pemetrexed were dominated by erlotinib. We found indications of superiority in terms of cost effectiveness for gemcitabine+cisplatin in a first-line setting, and for erlotinib in a second-line setting.