Abstract
This study compared the use of different test models to assess the cytotoxicity of a dental composite. The cytotoxicity of a composite polymerized using two halogen‐based light‐curing units (LCUs) (Max LC and Astralis) and two light‐emitting diode LCUs (E‐light and Freelight) served as the basis of comparison. Disk‐shaped specimens (7 mm diameter, 2 mm high) were fabricated using the four different light sources. The specimens were used in several cytotoxicity test models: direct and indirect contact tests as well as an extract test with an established cell line L‐929. The cells were stained with neutral red after cell‐material contact for 48 h. Neutral red‐stained areas (in mm^2^, for direct and indirect tests) and absorbance readings (for extract tests) were analysed statistically using ANOVA and the Tukey post hoc test, with P < 0.05 considered to be significantly different. Good correlation between direct and indirect contact tests (r = 0.903) was found. The extract test was the least correlated among the three tests. It was found that the E‐light + Freelight‐cured composite elicited cytotoxicity from the correlated studies. Uncured specimens were most detrimental to the cells in all tests. Our data demonstrated that composite cured with light‐emitting diode LCUs were cytotoxic to L‐929 cells. Different test models were found to give rise to different findings. Thus, a good cell‐material contact method would replicate more closely the physiological situation in vivo. This in turn would give more clinically relevant results. Copyright © 2005 John Wiley & Sons, Ltd.