One hundred and seven 5th-8th graders were tested on spatial rotation multiple-choice items to determine age and gender differences in spatial ability. Thirty-one of them were subsequently interviewed. They were asked to explain their reasoning when solving 4 of the tested items and a problem-solving task. Features of visual and non-visual strategies used were identified, however students did not make consistent use of only one type of strategy across tasks. They switched between visual and non-visual strategies or made combined use of both. Particular task characteristics may have influenced the choice of strategy in each case. Older students seemed to perform better on items involving three-dimensional manipulation but no other age and gender differences in strategy use and in spatial ability were detected. (Contains 27 references.) (Author/MM)