The present study evaluated the efficacy of electronic foramen locators (EFLs) to control root canal working length during rotary instrumentation and to assess possible reliability variations of different working lengths. Forty-eight human mandibular bicuspids were randomly divided in 2 groups according to the used device, Root ZX II (RZX) and Propex II (PRO). They were further subdivided in 2 subgroups according to the root canal preparation level (0.0 and -1.0). Preparation was performed with the Protaper rotary system using a crown-down technique. RZX was employed on its automatic auto-reverse mode (AAR) and PRO was used with the MPAS-10R contra-angle to monitor the preparation. The last used file (F3) was fixed, and the apical portion of the teeth was worn buccolingually, allowing to measure the extent between the file tip and the apical foramen (AF). The precision values of 0.0 mm and -1.0 mm were 100% and 0.0% for RZX, and 100% and 66.7% for PRO, respectively, with a range of ±0.5 mm. Statistical analysis showed no differences between the groups at 0.0 mm. However, at -1.0 mm, RZX showed the poorest results (0.96±0.11 mm), followed by PRO (0.43±0.23 mm). The difference between RZX and PRO was statistically significant. The EFLs were precise in maintaining the working length during rotary preparation when reaching the AF, but when their penetration was limited, both devices showed decreased precision; the RZX AAR failed in all instances.