Twenty children aged 3;0 to 3;10 were studied for behavior related to the acquisition of a single new word ("chromium," which was presented as designating the color olive green). The research was conducted in three cycles: prior to exposure to "chromium," at the time of a single encounter with that word, and about a week after the first encounter. On a sorting task, performance improved markedly from one cycle to the next. A comprehension task revealed that the children varied in their assignment of a particular color as referent for "chromium," with 47% giving a correct assignment at cycle 2 and 63% at cycle 3 (compared with 35% for a control group). A naming task revealed some evidence that experience with "chromium" influenced the child's naming category for olive. Of the children who understood a hyponym task, none changed his/her judgment whether chromium was a color between the second and third cycles. The failure of half the children to learn anything makes conclusions difficult but points to further considerations, including the cause of "fast mappings." (JB)