Evidence from Project STAR has suggested a considerable advantage of being in small classes in early grades. However, the extra benefits of additional years in small classes have not been discussed in detail. The present study examined the additional effects of being in small classes for more than 1 year. We find that once previous grade achievement and small-class membership are controlled for, small-class effects are typically insignificant except in 1st grade. Findings from longitudinal analyses showed no independent effects of small classes in early grades. Again, the exception was the estimate in 1st-grade mathematics. Thus, there is little evidence about the accumulated benefits of additional years in small classes. If indeed being in small classes for only 1 year is sufficient in early grades, the evidence points to reducing class size in first grade. (Contains 5 tables.)