AbstractThe present study tested whether the paradoxical technique of positive reframing, presented in the form of test instructions, could be used to reduce the cognitive experience of worry and thereby improve the arithmetic test performance of highly test anxious third‐and fourth‐grade children. Thirty high test anxious and thirty low test anxious children were assigned randomly to receive one of three types of instructions prior to taking an arithmetic word problem test: neutral instructions, reassuring instructions, and positive‐reframing instructions, which encouraged the children to view worry as a positive attribute. The results indicated that the type of instructions given had no significant effect upon subsequent test performance. However, both high and low test anxious children who received positive reframing scored significantly higher on a state anxiety measure given post test. These findings suggest that, although positive reframing may have an emotional impact, it is ineffective in improving the arithmetic test performance of highly test anxious chi