The identification, diagnosis, and categorisation of students who qualified for special education have created long-standing controversy. This article explores Maximilian P. E. Groszmann's measurement practices, which were intended to facilitate instruction in the early twentieth-century United States. In 1900, Groszmann established a private residential school for "atypical" children and developed his own method of measuring these children's capabilities, an alternative to the Binet scale. "Atypical" children were defined as those who deviated from the average human type but had the potential of a "normal" child. Thus, they were children who were classified between normal and abnormal. Groszmann's methods were based on his conviction that "atypical" children's development was closely and intricately related to their social environment. His holistic perception of "atypical" children's development led Groszmann to try to take a wider view of their difficulties, reactions, and experiences. Groszmann's answers to the questions of what, where, and why to measure children differed from those of advocates of intelligence testing. In short, his methods were intended to help children adapt to their social environment, not to label and sort them as "atypical" or socially incapable.