Second language (L2) acquisition researchers have acknowledged the need for an index of development (Hakuta 1975), comparable to the MLU (Mean Length of Utterance) of first language acquisition research, with which L2 researchers could compare populations with similar target language proficiencies. In an attempt to delineate a L2 index of development, 48 compositions written by non‐native speakers of English were examined. On an impressionistic basis two researchers independently assigned each composition to one of five levels of proficiency. The compositions at each proficiency level were then analyzed in order to identify the features that made each level unique. The features considered included: writing mechanics, clarity, organization, grammar, lexical choice, number of words, number of T‐units (Hunt 1965), average length of T‐units, number of error‐free T‐units, sentence construction, and content. The measures which seemed most suitable as a basis for an index of development were the average length of the T‐unit and the total number of error‐free T‐units per composition.Results, though incomplete, are encouraging. If the successful identification of stages in L2 development can be accomplished, much of the controversy resulting from conflicting claims about learning could be resolved since researchers would be making claims about learners at comparable stages of acquisition of the