Evaluators use their eyes to see what is there, whether it is intended or not. But they use their test instruments to measure what is intended, whether it is there or not. Evaluators have been broadening their repertoire of instruments for years: curriculum-embedded tests, observer checklists, audiotape recorders, videotape recorders, unobtrusive measures, the critical incident technique, situational tests, peer ratings, projective tests, criterion-referenced tests, and on and on. Developing tests in field situations promises to enrich their repertoire still further. In this article, the author discusses how to develop an instrument that can measure as well as one's eyes. He discusses three test developments that would produce hard evidence which would coincide with soft evidence: (1) field-based test development; (2) goal-free test development; and (3) theory-based test development.