Esophageal speech, although difficult to learn, is one of the ideal voice rehabilitation methods following total laryngectomy. The authors evaluated the possible factors in learning to use the esophageal voice and evaluated the intelligibility of esophageal speech by logopedic and phoniatric characteristics in 46 laryngectomees. High success rate(63.0%) was apparent in patients with wide residual pharyngeal mucosa(p<0.05), pharyngeal neurectomy(p>0.05), and success in air insufflation test(p<0.05). However, in cases of old age, in hypopharyngeal cancer patients and those who underwent neck dissection the success rates were poor, therefore a second choice of voice rehabilitation such as voice prosthesis is recommended. The most intelligible consonant in the manner of articulation was nasals(55.0%) and the least one was fricatives(22.8%)(p<0.05). However, the mean intelligibilities of consonant in the place of articulation were similar in various consonants. To improve the low intelligible consonants, esophageal speakers should practice increasing the volume of air inhalation, in addition to practicing the pronounciation.