Companies across the United States provide workplace English classes to non-native-English-speaking employees to increase productivity, retention, and on-the-job safety. Institutions that financially support the programs often require evidence of learning through standardized tests as a prerequisite for continued funding. However, the tests present challenges. It is often hard to find accepted tests that match the highly customized content and level of the classes. It is not clear if the learners can adequately demonstrate their learning as measured by standardized assessments, or if alternative assessments could contribute more meaningfully to their estimations of learning. In this study, 21 learners in a 10-week workplace English course took standardized pre- and posttests of listening and reading proficiency. The participants' learning was also measured qualitatively by learning journals. Paired correlations and a paired samples t test applied to the learners' scores on the standardized tests demonstrate that the learners did not perform significantly better after 30 hours of instruction. However, the learning journals demonstrated considerable learning gains, especially in terms of increased motivation, confidence, willingness to communicate, and job satisfaction. Discussed are reasons behind the discrepancies between the two forms of testing and implications for funding agencies and workplace English language programs alike. (Contains 6 footnotes and 7 tables.)