This paper presents an evaluation of Army job training development and testing practices, with a focus on Advanced Individual Testing. Information comes from intensive interviews with school instructors and from observations in the schools. Results indicate that some aspects of the Instructional Systems Development (ISD) model have been implemented less than ideally in Army training installations, while others have been implemented well or are improving. Many testing and training practices are not standardized within or across courses, though courses are almost all lock-step or group-paced at this time. Reliability and validity indices are not calculated by most schools. On the positive side, most tests appear to be criterion-referenced and hands-on, though there is a blurring of distinctions among test types in current Army use. Seven major implications are drawn from these findings. They relate to standardization of training development and test practices, interaction between schools which develop training and tests and schools which instruct, the use of audit trails, statistical properties of tests, the use of hands-on tests, philosophical and instructional conflicts, and the cyclic nature of Army training development and testing practices. (Author/BW)