We are comparing the cognitive load induced by various types of noise in the processing of information from speech. We examine if there is a difference in cognitive load between external noise sources (background noise) and internal noise sources (dysphonic voice). Our hypothesis is that noisy voices could be more cognitively demanding than background noise because they are more similar to the target signal spatially and temporally and the perceived link with the target signal is stronger. 60 normal hearing subjects (18-30 years) listen to texts in ten different conditions: (1) Healthy voice in multitalker babble noise, (2-4) Three types of dysphonic voices in background noise, (5-7) Three dysphonic voices in silence, (8) Healthy voice in silence, (9) Healthy voice with omnidirectional noise equivalent to dysphonic noise, (10) Healthy voice with dysphonic noise from same direction. We evaluate cognitive load in 4 ways: (1) during the listening phase by subjects’ performance on a secondary graphic task, (2) after the listening phase by subjects’ answers to a multi choice questionnaire and (3) by a free recall of the text, (4) by subjects’ grading of their listening effort. Methodological considerations will be discussed and preliminary results will be presented.