This resource paper is a guide for planning and developing instructionally relevant tests of student learning at the classroom, building, or district level. It is based on a model of instruction and testing which systematically uses assessment information to support and facilitate instructional improvement. Course goals and objectives are first translated into domain specifications which are used to link testing and instruction. A domain specification contains six components: (1) domain description; (2) content limits; (3) distractor limits and response-criteria; (4) item format; (5) student directions; and (6) sample items. Test items are then developed to match domain specifications. Standard item construction rules are given for both constructed responses (essay, short answer, completion) and selected responses true-false, matching, multiple choice). Items are then judged for their match to the six domain components plus linguistic and thinking complexity using the Item Rating Scales. When all eight categories have been individually scored on scales from 0 to 10, an overall rating is calculated. Guidelines are given for interpreting overall item rating for acceptance, rejection, or revision. Appendices contain: (1) sample domain-referenced test specifications; (2) item difficulty levels; and (3) materials used in the item rating process. (BS)