Objective: To evaluate the prognostic influence on local control and survival in vary patterns of nasal fossa involvement in nasopharyngeal carcinoma (NPC) after radiation therapy. Methods: Between November 1989 and July 1991, 218 patients with histologically diagnosed local-regional NPC were treated with radiotherapy. Based upon the fiberscope finding, we separated nasal fossa involvement into two types: exophytic protruding (EP), which indicated exophytic bulky tumor arising from the nasopharynx and protruding into the nasal fossa, and mucosal infiltration (MI), which indicated that the nasal cavity mucosa was clinically infiltrated by tumor. Of the 218 patients, 87 had nasal involvement. Sixty of them had a pattern of MI and another 27 had an EP component. Results: The likelihood of residual disease after irradiation, the local relapse rate, 5-year free from progression rate (FFP) and death rate associated with nasopharynx relapse (DRANP) of MI and EP were 36.7% vs 3.7%, 30.0% vs 7.4%, 26.7% vs 51.8% and 25.0% vs 3.7% with P<0.004, P<0.005, P<0.02 and P<0.03, respectively. Multivariate analysis in this selected group demonstrated that infiltration of nasal fossa mucosa was an independent prognostic factor on primary control and FFP. Conclusion: Differentiation of nasal fossa involvement according to MI or EP is of value in predicting the outcome of treatment. A specific biological difference between the MI and EP group might contribute to the statistical differences in the treatment endpoints we have observed. Objective: To evaluate the prognostic influence on local control and survival in vary patterns of nasal fossa involvement in nasopharyngeal carcinoma (NPC) after radiation therapy. Methods: Between November 1989 and July 1991, 218 patients with histologically diagnosed local-regional NPC were treated with Based on the fiberscope finding, we agreed nasal fossa involvement into two types: exophytic protruding (EP), which indicated exophytic bulky tumor arising from the nasopharynx and protruding into the nasal fossa, and mucosal infiltration (MI), which indicated that the Nasal mucosa was clinically infiltrated by tumor. Of the 218 patients, 87 had nasal involvement. Sixty of them had a pattern of MI and another 27 had an EP component. Results: The likelihood of residual disease after irradiation, the local relapse rate, 5-year free from progression rate (FFP) and death rate associated with nasopharynx relapse (DRANP) of MI and EP were 36.7% vs 3.7%, 30.0% vs 7.4%, 26.7% vs 51.8% and 25.0% vs 3.7% with P<0.004, P<0.005, P<0.02 and P<0.03, respectively. Multivariate analysis in this selected group demonstrated that infiltration of nasal fossa mucosa was an independent prognostic Factor on primary control and FFP. Conclusion: Differentiation of nasal fossa involvement according to MI or EP is of value in predicting the outcome of treatment. A specific biological difference between the MI and EP group might contribute to the Have observed.