A pilot study was conducted on test item ambiguity in standardized tests. A group of second and third grade children were selected and given a random sample of items from four standardized test series. Items were drawn from the reading, science and social studies subtests and put together in a composite reading comprehension, science, and social studies test. The children were interviewed individually about their perceptions of and answers to each item. The interviews were audiotape recorded and given analytical and holistic ratings. In both rating schemes, ambiguity in children's interactions with test items were of two types. The researchers attempted to distinguish three types of keyed answer ambiguity in both types of ratings; selection of a keyed answer by guessing, by mistake, or by other reasoning which did not reflect the skill measured. The limitations of the study were cited and explained. Ambiguity was seen as a real phenomenon, but substantially less of a problem than some critics indicated. The issues of ambiquity raised by this study refer more to validity than reliability. The authors suggest further research into children's perception and reasoning about tests and test items, as well as test interpretation and use. (DWH)