This paper presents the research findings of a study of the communicative competence of preschoolers. Twenty "language-impaired" and twenty normal speaking children, ages three to four, were observed and studied by the researcher using ethnographic methods. The ethnography of communication was presented by examining data gathered by using a "naturalistic/experiential" instrument which was designed based on the constructivist/holistic framework of studying language. The communicative competence of the children was studied in multiple contexts. Audio and visual recordings were used and the interactions were compared and analyzed qualitatively and quantitatively to see if there were significant differences between the two groups. The quantitative analyses included (1) t test, (2) logistic regression, (3) cluster analyses, and (4) correlation coefficient. The qualitative analyses included (1) range of variability, (2) generalizability of the communicative process, and (3) inferences about the nature of language impairment. The results of the quantitative analyses were helpful in identifying specific features of the language-impaired children's communicative patterns. However, they were not sufficient in providing a holistic picture of the children's abilities as well as difficulties. The results of the qualitative analyses indicated that the "traditional" approaches to assessment was too rigid and did not take into consideration individual differences. The results further revealed the complimentary nature of the quantitative and qualitative analyses. The research clearly indicated the need for more qualitatively structured and heuristically useful instruments for the diagnosis of language-impaired children. The study raised fundamental questions regarding the needs for institutions to hold a holistic view of language-impaired children, to assess language abilities and difficulties using an experiential/naturalistic approach and to respect individual differences. Furthermore, the research brought to light the complexity of the communicative process and the phenomenon of language impairment. The research indicated that considerably more work is required to understand the nature of the communicative process, to sharpen our analytical tools to evaluate children's repertoire, to improve observation methods, to avoid "pigeon-holing" children and to provide treatment and strategies when working with them. Finally, the paper made recommendations to teachers and clinicians for providing better diagnosis and interventions.