This study reports a secondary analysis from a randomized control trial study conducted with third graders with mathematics difficulty (MD; N = 75). We used simple linear regression to investigate whether students' performance on a word-problem vocabulary assessment predicted their performance on a measure of word-problem solving. Additionally, we used multiple linear regression to determine whether English learner (EL) status (n = 21; 28%) moderated this relation. Students' word-problem vocabulary knowledge significantly predicted their word-problem performance (R[superscript 2] = 0.27), but this relation was not moderated by EL status ([delta]R[superscript 2] = 0.003). These findings suggest word-problem vocabulary knowledge is important for word-problem solving for students with MD, regardless of EL status. Given the results showed that improved word-problem vocabulary was associated with higher levels of word-problem solving, researchers and teachers might consider including word-problem vocabulary support as part of word-problem interventions.