This study provides information on the onset of speech disfluencies in early childhood that are indicative to the development of stuttering and therefore explores their variations. It also provides an insight into the effectiveness of early childhood stuttering identification, i.e. the ways in which speech-language-pathology intervene on the matter. The purpose of the study is to provide helpful information that will be used by clinicians as a landmark in diagnostic procedures. Data were collected from the SLP's office in Center of rehabilitation - Faculty of Education and Rehabilitation Sciences. The sample included 47 children (26 boys and 21 girls) age ranging from 2,00 to 6,00 years. All children passed through differential diagnosis of speech disfluencies and their initial results are analyzed in this paper. The data are based on qualitative and quantitative analysis. Speech disfluencies most commonly occur in the child's second year of life. Their onset is earlier seen in girls than in boys. Most parents express concern about the observed onset and therefore seek SLP’s support shortly after recognizing speech disfluencies, which is also statistically confirmed. By comparing the frequency distribution of stuttered disfluencies, it was found that younger (2;00 to 4;00) and older (4;01 to 6;00) children do not differ in the frequency of their occurrence. Most children are at risk for stuttering. This information points to the need for systematic monitoring of these children. The analysis has shown that children with different assigned diagnosis do not differ in the form of involvement in therapy. Most children are involved in indirect therapy. Indirect and direct forms of involvement also include children with diagnosis of stuttering that makes the most of the observed sample (57.4%). Due to the fact that older children are more at risk for developing chronic stuttering, they are more often involved in direct therapy. In the process of making diagnostic decisions, the clinician is often carried out by his subjective judgments because of non existing standardized measuring instruments on Croatian language. Inconsistent use of diagnostic terminology and unsystematic recording of diagnostic records have been noted. Current state points to the problems in the domain of diagnostics of early childhood stuttering that require changes.