This paper presents the application of Cox regression model in survival analysis of lung cancer for the first time in China.By using Cox model

a multivariate analysis was made on 191 patients with lung cancer treated by operation. Among the twenty three factors

only pathological sort

the size of tumor

lymph node metastases

type of operai ions

infiltra -ing across the end of tumor

and complications showed a high significance with prognostic relationship. Of those factors influecing postoperative survivaltime

lymph node metastases

is detrimental for patients

squamasum carcinoma and lobectomy are preferable. From survival curves of groups of different harzad groups

fit in the model and survival rale in the interclass was calculated

Therefore it showed advantages using Cox regression model in multivariate survival analysis