Introduction: The aim of this in vitro study was to evaluate the effect of smear layer on the push-out bond strength of white mineral trioxide aggregate (WMTA) with and without disodium hydrogen phosphate (Na2HPO4). Materials and Methods: Dentin discs with standard cavities were obtained from extracted human single-rooted teeth and divided to 4 groups (n=15) according to the irrigation regimen and the canal filling material. In groups 1 and 3, canals were irrigated with normal saline; in groups 2 and 4, irrigation method included sodium hypochlorite (NaOCl) and then ethylenediaminetetra-acetic acid (EDTA). The canals were filled with WMTA in first and second groups and with WMTA+Na2HPO4; in groups 3 and 4. The samples were wrapped in wet gauze and incubated in 37°C for 3 days. The push-out bond strength was then measured by means of the Universal Testing Machine and the failure modes were examined under stereomicroscope at 40× magnification. Tow-way ANOVA was used to evaluate the effect of material type and smear layer removal. Post hoc Tukey test was used for the two-by-two comparison of the groups. Results: The greatest and lowest mean±standard deviation for push-out bond strength were observed in groups 4 (4.54±1.14 MPa) and 1 (1.44±0.96 MPa), respectively. The effect of removing the smear layer on the push-out bond strength of WMTA+Na2HPO4 was significant (P=0.01), but not for WMTA (P=0.52). Interestingly, there was significant difference between groups 1, 3 and 2, 4 (P<0.05). The failure mode for all experimental groups was of mixed type. Conclusion: Under circumstances of this in vitro study, removal of smear layer increases push-out bond strength when Na2HPO4 is added to WMTA.