Previous studies have shown that iconic gestures can be integrated with speech for better comprehension in both first and second language contexts. Yet it remains unclear whether such speech-gesture integration is found in consecutive interpreting (CI) and influenced by language tasks (passive viewing or CI) and interpreting experience (individuals with or without CI experience). To this end, we conducted two experiments through a modified cued-recall task in which Chinese-English bilinguals were exposed to utterances accompanied by three gesture conditions: semantically related gesture, semantically unrelated gesture, and no gesture. In Experiment 1, bilinguals with no CI experience performed the cued-recall task after either a passive viewing task or a CI task. In Experiment 2, bilinguals with CI experience only performed CI tasks before the task. Experiment 1 showed that bilinguals responded faster and more accurately to semantically related gestures than other gesture conditions in both language tasks. The results of Experiment 2, compared with the performance of bilinguals in the CI task in Experiment 1, also revealed the facilitative role of iconic gestures for language comprehension during CI. However, the two experiments found no interaction effect between gesture condition and language task or CI experience, suggesting speech-gesture integration is not affected by language task and CI experience. This strengthens the robustness of the integrated-systems hypothesis that people integrate speech and gestures for L2 comprehension even in CI activity.