Classification of laryngeal dysplasia, the most appropriate treatments and criteria for evaluation of the results is still a highly controversial issue. The objectives of the present study on the treatment of laryngeal dysplasia lesions are to: 1) evaluate the relative incidence of the various forms of dysplasia in relation to grading of the histopathological findings; 2) establish the prognosis for the various forms of dysplasia considered; 3) determine the results achieved by the author's treatment protocol according to the characteristics of the dysplasia; 4) critically evaluate the classifications of laryngeal dysplasia found in the literature in view of the results of the present study. The study involved 141 patients with vocal cord dysplasia (134 men, 7 women; mean age: 56.2 years) who had come under observation at the E.N.T. Dept. of the University of Naples "Federico II" between January 1981 and April 1998. In all cases the dysplasia was removed by CO2 laser microlaryngoscopy. Of the 141 patients 89 (63.2%) showed mild dysplasia, 14 (9.9%) moderate dysplasia, 20 (14.2%) severe dysplasia and 18 (12.7%) in situ carcinoma. The five-year survival rate showed an overall actuarial survival of 89.1% for all patients while the corrected actuarial survival was 98.5% and local disease control was 86.1%. In 17 cases (12%) the dysplasia lesion recurred, in 11 (7.8%) an infiltrating carcinoma arose. Recurrences in the dysplasia were encountered in 9% of the patients with mild lesions, 7.1% of those with moderate dysplasia, 15% of the subjects with the severe form and in 27.7% of those with in situ carcinoma. An infiltrating carcinoma arose in 5.6% of the cases of mild lesion, in 7.1% of the medium dysplasias, 5% of the severe forms and in 22.2% of those with in situ carcinoma. The recurrences and infiltrating carcinomas were successfully treated with endoscopic CO2 laser surgery. Only three cases (2.1%) required radical surgery (total laryngectomy): these were patients who had not quit smoking and who had not adhered to the planned follow-up. One of the latter patients died with widespread metastases of the laryngeal cancer. The significantly higher incidence of dysplasia recurrences (P = 0.028) and infiltrating carcinomas (P = 0.015) in those patients with in situ carcinoma shows that, when preparing classification and determining prognosis, these pathologies should be considered separately. Moreover, they require a particularly precise follow-up. In conclusion, the following observations can be drawn from the study: mild and moderate dysplasias are more frequent as compared to severe dysplasia and carcinoma in situ; carcinoma in situ should be distinguished for classification and prognosis from dysplasias, considering their different clinical course and evolution; CO2 laser surgery in these forms is a valid therapeutic approach for its precision, the achievable results, the reduced discomfort to patients and for its evident cost-effectiveness.