In this compartive study, we were mainly concerned with questions, such as "What kind of difficulties in lip-reading are overcome by speech-hearing under the best condition of the use of residual hearing?" "Is the basic capacity in speech-perception fully utilized or not?" and "How many erroneous responses in speech-reception can be classified in a sytematic order of approximation to speech stimuli?". The eight adolescents who took part in our two experiments had hearing losses over sixty db at several frequencies of pure tones. Among them, the subject "A" was slightly deafened in her early school age, and had near-total loss of hearing in her adolescence. Although the subject "B" had been deafened severely before his infancy, his hearing losses were not over seventy db at a recent date. Both of A and B had good reading abilities and normal intelligence. Their lip-reading abilities were above average. In this paper the experimental data of these two subjects are analyzed in detail. In our first experiment, 542 speech-stimuli (including words, phases, short incomplete sentences and small numbers of nonsensical syllables) were arranged in 476 pairs of speech-stimuli. Each subject was asked (1) to judge whether the two speech-stimuli of each pair were perceived as same stimuli or not, (2) to record the speech by syllable-lettters, (3) to make a confidence-rating in an attempt to estimate his or her accuracy of response. In the second experiment, the relations between erroneous responses to a few speech-stimuli were followed up with successive four experiments of speech reception. As to the subject A and B, we infer : 1. Their difficulties in speech-reception lie in : (a) the same words were frequently judged to be different words, (b) the chain of erroneous responses spread to a considerable extent, (c) expressions of self-confidence to the responses were not successful, (d) the correct reception of nonsensical syllables was impossible. 2. In a training of their speech-reception, it would be important to note : (a) the pairs of speech-stimuli were highly discriminmated by speech-hearing, (b) many erroneous responses by lip-reading follow the rule-of-homophenous-response, (c) there were characteristic differences in quality of erroneous response between speech-hearing and lip-reading, (d) many different kinds of difficulties in lip-reading were sufficiently overcome by speech-hearing ; however, lip-reading also had some advantages, (e) there was a qualitative difference between subject A and Subject B.