A brief survey by questionnaire was conducted to show that the readability of newspaper articles on science depends on individual articles. In other words, some kind of articles might generally be considered difficult to read. An open-ended question gave the clue to what gives readers the impression that some stories are difficult and others easy. Some specific words related to the difficulty were extracted from the responses, such as knowledge, interest, specialty, words etc. It leads to the presumption that people tend to think that scientific articles are easy to understand when they contain fewer technical terms and require less previous knowledge of the field they deal with. It was also indicated that readers feel less difficulty in reading a certain article when similar news are often reported by newspapers or TV and have become familiar to them. These results have not been specified in previous studies in Japan, and will be quite helpful in finding the way that the society members share an increasing amount of scientific information these days.