When spoken information is presented in unfavourable listening conditions a larger part of the available and limited information processing resources must be allocated to the identification of the words spoken. This leaves fewer resources for the further processing of the speech (the interpretation and storing of the information). In such circumstances understanding and memory of speech might be impaired although each word has been correctly heard. This hypothesis was confirmed in two previous studies where memory of a word list was shown to be impaired when presented in unfavourable listening conditions: unfavourable signal/noise ratio (Kjellberg, Ljung & Hallman, 2008) and a too long reverberation time (Ljung & Kjellberg, in press). Kjellberg et al. also found that the noise effect was weaker the better the working memory capacity (as measured with a reading span task).