Abstract Alternative photoinitiators with different absorption wavelengths have been used in resin composites (RCs), so it is crucial to evaluate the effectiveness of light-curing units (LCUs) on these products. Objective Using Fourier transform infrared analysis (FTIR) in vitro, the effects of varying radiant exposure (RE) values generated by second and third generation LED LCUs on the degree of conversion (DC) and maximum rate of polymerization (Rpmax) of an experimental Lucirin TPO-based RC were evaluated. Material and Methods 1 mm or 2 mm thick silicon molds were positioned on a horizontal attenuated total reflectance (ATR) unit attached to an infrared spectroscope. The RC was inserted into the molds and exposed to varying REs (18, 36 and 56 J/cm2) using second (Radii Plus, SDI) and third generation LED LCUs (Bluephase G2/Ivoclar Vivadent) or a quartz tungsten based LCU (Optilux 501/SDS Kerr). FTIR spectra (n=7) were recorded for 10 min (1 spectrum/s, 16 scans/spectrum, resolution 4 cm-1) immediately after their application to the ATR. The DC was calculated using standard techniques for observing changes in aliphatic to aromatic peak ratios both prior to, and 10 min after curing, as well as during each 1 second interval. DC and Rpmax data were analyzed using 3-way ANOVA and Tukey’s post-hoc test (p=0.05). Results No significant difference in DC or Rpmax was observed between the 1 mm or 2 mm thick specimens when RE values were delivered by Optilux 501 or when the 1 mm thick composites were exposed to light emitted by Bluephase G2, which in turn promoted a lower DC when 18 J/cm2 (13 s) were delivered to the 2 mm thick specimens. Radii Plus promoted DC and Rpmax values close to zero under most conditions, while the delivery of 56 J/cm2 (40 s) resulted in low DC values. Conclusions The third generation LCU provided an optimal polymerization of Lucirin TPO-based RC under most tested conditions, whereas the second generation LED-curing unit was useless regardless of the RE.