AbstractThe current 2‐part study investigated the relative contributions of English language training (ELT) and computerized Cogmed working memory training (WMT) to improvements in English‐as‐a‐foreign language (EFL) proficiency and working memory capacity. In Study 1, Japanese undergraduate EFL learners were randomly allocated to 1 of 3 experimental groups (ELT only, WMT only, and both WMT and ELT) and a control group. The dual intervention (WMT + ELT) group, who received 5 weeks of both WMT and ELT, retained the training effects on verbal short‐term memory and working memory at 6 months post‐training. No significant gains were made, however, in a standardized EFL test of vocabulary, grammar, reading, and listening comprehension abilities. Addressing the absence of a measure of more productive skills in Study 1, Study 2 was carried out on a new cohort of Japanese undergraduate EFL learners comprising a WMT group and a control group. No significant far‐transfer effects (on the language‐based measures) were identified. Descriptive and correlational analyses revealed that the higher general oral proficiency scores tended to be associated with larger improvements in a verbal working memory task (involving storage and manipulation of letter sequences) completed during the Cogmed WMT program. These analyses indicate the possibility of more specialized WMT programs translating into more tangible gains in second or foreign language performance. Further insights into the limited far‐transfer effects of WMT could be gleaned by exploring a modular approach to studying working memory.