OBJECTIVE: The aim of this study was to establish if dental panoramic tomographic (DPT) radiographs provide a reliable means of assessing the anterior maxilla in new orthodontic patients. STUDY DESIGN: Two hundred fifty pairs of upper anterior occlusal (UAO) and DPT radiographs were randomly selected from the records of new orthodontic patients. Ten observers each rated 50 pairs of radiographs for specific risk factors as well as for a measure of adequacy. Correlations between the 2 radiograph types and inter/intraobserver reliability were computed using Cohen kappa test. RESULTS: Intraobserver reliability was good for detection of impacted teeth/supernumeraries/periapical pathology (kappa = 0.711-1.0). Reliability was more variable for the specific root morphologies, with blunt and bent roots being reassessed as such most reliably (kappa = 0.259-0.533). There was low interobserver reliability (kappa = 0.327 for UAO and kappa = 0.223 for DPT) for "normal" or "not normal" root morphology. For every variable, however, the interexaminer agreement was higher when using UAO radiographs rather than DPTs. Observers were best able to agree on the diagnosis of eroded/resorbed roots on the UAO (kappa = 0.402) and blunt roots (kappa = 0.303) on the DPT radiographs. With UAO as the gold (reference) standard for grading root morphology, DPT had a sensitivity of 45.6% and a specificity of 71.4% for detecting abnormal root form. Risk factors were more likely to be detected on the UAO radiographs than the DPTs. DPTs had a poor ability to detect abnormal root form. DPTs were more likely to be rated as "inadequate" than UAO radiographs. On 6 occasions, supernumerary teeth that were observed on the UAO were missed on the DPT. CONCLUSION: The DPT is not an accurate means of screening the anterior maxilla prior to orthodontic treatment.