This study uses techniques of experimental syntax to explore island effects and the that-trace effect in three populations: Early arrival Korean immigrants to the U.S. (AoA=6-10), late arrival Korean immigrants (AoA=12-15), and L1 English controls, all with a length of residence in the U.S. of at least 7 years. It is seen that all three groups show island effects, but only the native controls show a clear that-trace effect. These results are analyzed in terms of a processing account of island constraints and a grammatical account of the that-trace effect based on the Subject Criterion. The results do not support the classical notion of a critical period and suggest that principles and mechanisms within syntax may be affected differentially by age-related factors.