Previous studies on various marine mollusc species have shown that both larval and juvenile growth rates are substantially heritable, but few workers have examined the extent to which larval and juvenile growth rates covary. We examined the relationship between larval and juvenile growth rates in seven laboratory experiments conducted between 1986 and 1993, using the prosobranch gastropods Crepidula plana Say and C. fornicata (L.). In most experiments larvae were reared individually, measured twice nondestructively to determine larval growth rate, allowed or stimulated (daily 5-h exposure to 20 mM excess K + in seawater) to metamorphose, and then measured at least twice after metamorphosis to determine juvenile growth rates. Generally, there was no significant (p > 0.10) relationship between larval and juvenile growth rates, suggesting that in these two species selection can act independently on the two stages of development. A positive correlation (p = 0.007) between larval and juvenile growth rates was observed for C. fornicata in one experiment, but only for offspring from females maturing the most rapidly in laboratory culture. Even for these larvae, however, variation in larval growth rate explained <2% of the variation in juve-Communicated by J.P. Grassle, New Brunswick J.A. Pechenik ([