Hilsa, Tenulosa ilisha is an important commercial fish of the Indo-Pacific region. The larvae of hilsa can survive for quite long without exogenous feed due to the presence of large yolk sac and oil globule. Therefore it is important to understand the suitable age of hilsa larvae to be used for nursery rearing. In this study, larvae were produced through artificial breeding of mature migratory fish collected from wild. Three different age groups of hilsa larvae, such as three days post hatched (3 DPH, just before yolk sac exhaustion, swim mostly vertically, TL: 4.21 ± 0.02 mm), 5 DPH (with oil globule, swim vertical and horizontal, TL: 5.26 ± 0.12 mm) and 12 DPH (just before oil globule exhaustion, swims very fast, TL: 6.00 ± 0.11 mm) were randomly distributed in nine well prepared rectangular earthen ponds (100 m2) in triplicate (n = 3) at the stocking density of 500 number m−3. The nursery rearing was carried out for 90 days. Results showed no significant difference in water quality parameters in different experimental ponds (p > 0.05). Dissolved oxygen, temperature, pH, total ammonia nitrogen, chlorophyll-a, gross primary productivity, total plankton count ranged from 7.24–7.46 mg L−1, 29.00–30.38 °C, 8.16–8.43, 110–124 μg L−1, 17.33–22.66 μg L−1, 193–203 mg C m−3 h−1, 2.2 × 104–2.4 × 104 numbers mL−1, respectively. After 60 days of the nursery, to imitate the larval migration towards low saline water, salinity was increased at the rate of 0.5 g L−1 on alternate days to 7.4 g L−1. There was no significant difference (P > 0.05) in daily weight gain and final total length and weight of fish in different experimental groups, however specific growth rate varied significantly (p < 0.05). At the end, hilsa larvae attained a total length of 43.3 ± 0.074 mm, 41.4 ± 0.04 mm, 39.2 ± 0.051 mm, in 3, 5 and 12 DPH groups, respectively. Percentage survival was significantly (p < 0.05) highest (3.40 ± 0.81%) in 5 DPH treatment. Water and gut plankton analysis showed that hilsa larvae are not a selective feeder, but feeds on available food items in the ecosystem through filtration. Length-weight relationship indicated the negative allometric growth (b < 3) pattern in all three groups. Overall, this experiment suggested that 5 DPH is an ideal age for shifting hilsa larvae from larval rearing tank to nursery pond.