This article examines the issue of the quality of teacher-produced tests, limiting itself in the current context to objective, multiple-choice tests. The article investigates a short, two-part 20-item English language test. After a brief overview of the key test qualities of reliability and validity, the article examines the two subtests in terms of test and item quality, using standard classical test statistics. Unsurprisingly, the pretested items outperform the teacher-produced test. The differences between the two subtests underscore issues about the quality (or lack thereof) of teacher-produced tests. The article ends with suggestions of how teacher-produced tests might be improved.