To evaluate and document the effectiveness of miner training, a study monitored a wide variety of classroom sessions conducted for new miner and annual refresher training. Seven trainers and 275 mine trainees at fourteen sites in Kentucky, Maryland, Pennsylvania, and West Virginia participated. Observations were made of classes in underground new miner and annual refresher training and in surface annual refresher training. Course materials were both instructional modules developed by the Bureau of Mines and instructor materials. Field data collection involved audiotaping of classroom sessions, pre- and post-testing of trainees, and interviews with students and instructors. Test scores alone showed significant gains. A one-way analysis of variance was performed on variables identified by expert opinion. As regards underground inexperienced training, results indicated that training eliminated educational differences seen in pretest analysis and type of training (company, freelance, vocational-technical, extension) was significant for both pre- and post-test. As regards annual refresher training, results indicated any previous training was a significant variable for pretest scores and years of mining experience were not significant. Measurement techniques frequently used in classic classroom investigations were found inappropriate for evaluating mine training, since miners had no motivation to perform well on tests. (YLB)