We report an outcome study of persons with aphasia participating in community-based treatment programmes. Patients (n=50) were assessed before and after treatment using: (i) a standardized test of impairment, the Western Aphasia Battery, administered by treating clinicians; and (ii) a standardized assessment of disability (functional communication), the Communicative Effectiveness Index, rated by family members. Pretreatment and postreatment means were calculated and compared, with matched t-tests utilised to probe statistical significance of improvements after treatment. We then calculate impairment- and functional-level means by aphasia diagnostic categories, assigning rank orders and calculating Spearman rank-order correlations. Data analysis shows that, before treatment, patients spanned a wide range of times after onset, aphasia diagnostic types, and severity levels at start of care. Following treatment, means of the 50 patients improved significantly on every measure administered at both the impairment and the functional levels. Absolute improvements ranged from 6.5-26.2%, with statistical significance ranging from p<.01 to p<<.001. Before treatment, there is strong positive correlation (p=+.90) between impairment-level and functional-level assessment means by diagnostic categories; after treatment, improvement means by these diagnostic categories show moderate negative correlation (p =-.60). Further examination shows that post-treatment improvements are found to be best viewed as functions of same-type severity levels of pretreatment, with patterns of improvement at the impairment and functional levels diverging distinctly.