The objectives of this study were threefold: (1) to evaluate the accuracy of clinical evaluation compared with videofluoroscopic swallowing studies (VFSSs) in the detection of penetration and aspiration in children of age 0 to 15 years presenting with feeding and swallowing problems; (2) to assess the relationship between therapists' confidence ratings in making judgements about the presence or absence of penetration and aspiration, and the accuracy of their evaluation as confirmed by VFSSs; (3) to identify clinical predictors of penetration and aspiration during clinical evaluation of children with feeding and swallowing difficulties. We used a prospective study to evaluate the sensitivity, specificity, and positive and negative predictive values of a diagnostic clinical evaluation compared with VFSSs (criterion standard). Clinical evaluation and videofluoroscopy forms for oral motor and swallowing evaluation, which included potential indicators of aspiration, were designed for this project. Seventy-five children with feeding problems participated (33 females, 42 males; age range 0 to 14 years, mean 2 years; 62% of participants younger than 12 months). For fluids, clinical evaluation showed a sensitivity of 92% for aspiration. For solids, sensitivity for detecting aspiration was 33%. Analysis of the therapists' mean confidence ratings compared with the accuracy of their judgement demonstrated that when therapists were very sure that the child was aspirating or penetrating or not, they were correct. When the therapists were unsure, then the accuracy of prediction was not as good. Cough was the most significant predictor (p < 0.05) of fluid aspiration and penetration. We conclude that clinical evaluation with experienced clinicians can detect aspiration and penetration of fluids in children of varied ages and diagnoses, but that it is not accurate with solids.