Objective To assess and compare the sensitivity and specificity of ultrasonography and magnetic resonance imaging in the diagnosis of placenta accreta in patients with placenta previa. Methods This retrospective cohort study included 37 women, and was conducted between January 2013 and October 2015; 16 out of the 37 women suffered from placenta accreta. Histopathology was considered the gold standard for the diagnosis of placenta accreta; in its absence, a description of the intraoperative findings was used. The associations among the variables were investigated using the Pearson chi-squared test and the Mann-Whitney U-test. Results The mean age of the patients was 31.8 ± 7.3 years, the mean number of pregnancies was 2.8 ± 1.1, the mean number of births was 1.4 ± 0.7, and the mean number of previous cesarean sections was 1.2 ± 0.8. Patients with placenta accreta had a higher frequency of history of cesarean section than those without it (63.6% versus 36.4% respectively; p < 0.001). The mean gestational age at birth among women diagnosed with placenta previa accreta was 35.4 ± 1.1 weeks. The mean birth weight was 2,635.9 ± 374.1 g. The sensitivity of the ultrasound was 87.5%, with a positive predictive value (PPV) of 65.1%, and a negative predictive value (NPV) of 75.0%. The sensitivity of the magnetic resonance imaging was 92.9%, with a PPV of 76.5%, and a NPV of 75.0%. The kappa coefficient of agreement between the 2 tests was 0.69 (95% confidence interval [95%CI]: (0.26–1.00). Conclusion The ultrasound and the magnetic resonance imaging showed similar sensitivity and specificity for the diagnosis of placenta accreta.