This study evaluated the influence of retorque on loosening torque (Lt) of prosthetic screws in implant-supported dentures with different fit levels. Ten mandibular implant-supported dentures were fabricated and then 20 cast models were then prepared using prosthetic structures to create 2 fit levels: passive fit (Pf) and misfit (Mf). Two tightening techniques were also evaluated: initial torque only (T1); and initial torque and retorque after 10 min (T2). Gold or titanium screws were used, resulting in 4 groups to each one: Pf/T1, Pf/T2, Mf/T1, Mf/T2. The Lt was measured 24 h after the tightening torque using digital torque meter. Data were analyzed statistically by two-way ANOVA and Tukey's test (alpha=0.05), separately for each screw material. For titanium screws, no significant difference (p>0.05) was found between Pf/T1 and Pf/T2, or between Pf/T2 and Mf/T2. However, statistically significant difference (p<0.05) was found between Pf/T1 and Mf/T1, and between Mf/T1 and Mf/T2. Mf reduced the Lt using T1, while and T2 increased the Lt for Mf. Retorque and fit were shown to have no significant influence on the Lt of the gold screws. Retorque application made insignificant the misfit effect on the Lt of the titanium screws, suggesting that this procedure should be performed routinely during the screw tightening in multi-unit dentures.