Many reports exist regarding the economic evaluation of evolving chemotherapeutic regimens or diagnostic images for lung cancer (LC) patients. However, it is not clear whether clinical information, such as pathological diagnosis or cancer stage, should be considered as a risk adjustment in lung cancer. This study compared the cost and practice patterns between small cell lung carcinoma (SCLC) and non-small cell lung carcinoma (NSCLC) patients. 6,060 LC patients treated at 58 academic hospitals and 14,507 at 257 community hospitals were analyzed. Study variables included demographic variables, comorbid status, cancer stage, use of imaging and surgical procedures, type of adjuvant therapy (chemotherapy, radiation or chemo-radiation), use of ten chemotherapeutic agents, length of stay (LOS), and total charges (TC; US$1 = 100 yen) in SCLC and NSCLC patients. The impact of pathological diagnosis on LOS and TC was investigated using multivariate analysis. We identified 3,571 SCLC and 16,996 NSCLC patients. The proportion of demographic and practice-process variables differed significantly between SCLC and NSCLC patients, including diagnostic imaging, adjuvant therapy and surgical procedures. Median LOS and TC were 20 days and US$6,015 for SCLC and 18 days and US$6,993 for NSCLC patients, respectively (p < 0.001 for each variable). Regression analysis revealed that pathological diagnosis was not correlated with TC. Physicians should acknowledge that pathological diagnosis dose not accounts for any variation in cost of LC patients but that should remain as an indicator of appropriate care like selection of chemotherapeutic agents.