This paper is concerned with soft computing techniques for categorizing laryngeal disorders based on information extracted from an image of patient's vocal folds, a voice signal, and questionnaire data. Multiple feature sets are used to characterize images and voice signals. A committee of support vector machines (SVM) is designed for categorizing the data represented by the multiple feature sets into the healthy, nodular and diffuse classes. The feature selection and classifier design is combined into the same learning process based on genetic search. When testing the developed tools on the set of data collected from 240 patients, the classification accuracy of over 98.0% was obtained. Combination of the three modalities allowed to substantially improve the classification accuracy if compared to the highest accuracy obtained from a single modality.