One of the major goals in mathematics education is to teach the ability to solve word problems. This study compared the word problem solving abilities of second-grade students from Taiwan (N=24) with those of American students (N=24). Two tests were used in the problem solving interviews. The comparisons between the performance of the groups on different problem solving subprocesses revealed that children from Taiwan were better than their American counterparts in schematic knowledge only on word problems using addition number sentences with unknown results. The groups did not differ in their semantic or procedural knowledge. Findings suggest that successful word problem solving involves an integration of different kinds of knowledge underlying word problem solving processes. (ASK)