The shrinkage stress continues to be the major drawback of restorative composites (RC). Objectives: This study evaluated the effect of the photoinitiator type on the maximum rate of polymerization (Rpmax), rate of shrinkage stress development (Rstressmax), degree of conversion (DC) and crosslink density (CLD). The hypothesis tested was that alternative photoinitiators could reduce the Rpmax and the Rstressmax while producing similar DC and CLD as camphorquinone (CQ). Methods: Experimental RC containing CQ (control), phenylpropanedione (PPD) or CQ/PPD were cured by halogen light, 40s. Rpmax was evaluated through differential scanning calorimetry (DSC). A cantilever-device evaluated the stress development during five minutes. The samples (disk shape, 10 mm Ø; 0.8 mm thickness) were subsequently tested by FTIR to evaluate the DC on two sides (top, close the light guide, and bottom). After ethanol storage, the swelling coefficient (α) was measured as an indirect method to evaluate the CLD. Data was submitted to ANOVA and Tukey's test (p=0.05). Results: Data analyses showed that: Rpmax (%/sec), CQ (3.3±0.1) >CQ/PPD (3.0±0.1) >PPD (2.4±0.0); Rstressmax (MPa), CQ (5.75±1.18) >CQ/PPD (4.09±0.36) >PPD (2.79±0.80); DC (%) top, CQ (50.51±3.09) = CQ/PPD (49.43±5.37) = PPD (47.67±3.81); bottom, CQ (52.60±3.37) = CQ/PPD (52.59±4.05) >PPD (46.16±1.87); α (q/ml), CQ (0.004±0.001) = CQ/PPD (0.005±0.002) = PPD (0.008±0.002) but CQ>PPD. Conclusion: CQ/PPD reduced the Rpmax and Rstressmax without reduction in DC and CLD.