Spat of blacklip pearl oyster, Pinctada margaritifera (L.), recently have been found to suffer mass mortalities every autumn in farms in subtropical Japan. We investigated what rearing conditions improve growth and survival of the spat in one of the farms. From September 2002, spat with mean ±SD dorsoventral measurement 14.5 ± 2.4 mm were reared in trays (33 cm × 21 cm × 8 cm) with combinations of different spat densities (12, 24, 36, 72 or 144 spat tray⁻¹), tray partitioning (trays with or without partitions) and suspension depths (2 or 6 m). Three months later, spat densities and tray partitioning significantly affected the growth: greater for 12–36 spat tray⁻¹ (mean ± SD: 12.2 ± 3.5 mm) than for 72–144 spat tray⁻¹ (7.7 ± 4.3 mm) and for partitioned trays (12.2 ± 4.1 mm) than for nonpartitioned trays (8.5 ± 4.0 mm). Yet, no set of rearing conditions significantly explained the great variation in the survival rate (mean: 51.4%, range: 0.0% to 100.0%). When two groups of spat reared separately in the farm and laboratory were transferred to the same aquarium, cross infection occurred. It is possible that spat-density and tray-partitioning dependent factors (e.g., food abundance and space per spat) affected mainly the growth, whereas other factors including a putative infectious agent affected mainly the survival rate.