The tri-spine horseshoe crab, Tachypleus tridentatus, is an endangered species in China, and artificial propagation followed by stock enhancement has become an important approach for resource recovery and conservation. The survival and developmental performance of hatchery-reared juveniles are crucial to the success of these practices. Among early life stages, 2nd-instar juveniles are at a key developmental period during the transition from dependence on endogenous reserves to exogenous feeding, and their tolerance to starvation may have important effects on subsequent growth, molting, and survival. In natural intertidal habitats, prey resources are unevenly distributed and feeding opportunities are restricted by tidal rhythms; therefore, early juveniles are likely to encounter intermittent food limitation. However, information remains limited on the starvation tolerance of already feeding 2nd-instar juveniles and on the extent to which post-starvation re-feeding can compensate for the adverse effects of food deprivation. To address this issue, the present study investigated the effects of different starvation durations followed by re-feeding on the growth, molting, and survival of artificially propagated 2nd-instar juveniles of T. tridentatus, with the aim of providing a scientific basis for hatchery management and stock enhancement. Artificially propagated 2nd-instar juveniles that had already initiated feeding were used in the experiment. Juveniles had molted to the 2nd instar approximately 2 weeks before the trial and had been continuously fed freshly hatched Artemia nauplii prior to the experiment. At the beginning of the trial, the juveniles had a mean prosomal width of 8.310 ± 0.326 mm and a mean body mass of 51.90 ± 2.70 mg. Eight experimental groups were established, including starvation for 0, 3, 6, 9, 12, 15, and 18 d, as well as a continuous-starvation group. Each treatment had three replicates, with 15 juveniles per replicate. Experimental conditions were maintained at 28 ± 0.5 °C, salinity 28-30, and dissolved oxygen ≥ 5.0 mg/L. During the starvation period, body mass loss, mortality, and molting were recorded. After starvation, juveniles in the 0-18 d groups were re-fed for 70 d. Feeding recovery was assessed 3 h after re-feeding, and final body mass, specific growth rate, survival rate, developmental stage, molting rate, and mean molting time were determined during the recovery period. The continuous-starvation group was used to evaluate the upper limit of starvation tolerance. Statistical analysis was conducted using one-way ANOVA followed by multiple comparison tests, with significance set at P < 0.05. The results showed that during the 0-18 d starvation period, body mass loss increased significantly with increasing starvation duration. The reduction in body mass remained low in the short-term starvation groups, but increased markedly after prolonged deprivation, reaching 15.65 ± 0.94 % in the 18 d group (P < 0.05). No mortality or molting occurred in any group during the first 18 d of starvation, indicating that already feeding 2nd-instar juveniles had relatively strong tolerance to short-term food deprivation. In the continuous-starvation group, mortality began on day 51, the median lethal time occurred on day 81, and all juveniles died by day 107, suggesting that prolonged starvation eventually exceeded the physiological tolerance of this developmental stage. After food was restored, juveniles in all re-feeding groups resumed feeding rapidly. The feeding rate measured 3 h after re-feeding ranged from 93% to 100%, with no significant differences among groups (P > 0.05). After 70 d of re-feeding, survival remained high, ranging from 80% to 96%, and no significant differences were observed among treatments (P > 0.05). Most juveniles developed to the 3rd or 4th instar by the end of the experiment. However, prolonged starvation caused clear carry-over effects on subsequent growth and development. In the 18 d starvation group, a small number of juveniles remained at the 2nd instar after 70 d of re-feeding, and no individuals reached the 4th instar. Compared with the 0 d group, final body mass and specific growth rate in the 9-18 d starvation groups decreased progressively with increasing starvation duration. In addition, the mean time to molt to the 3rd instar was significantly prolonged, increasing from 24.78 ± 0.14 d in the control group to 45.41 ± 0.15 d in the 18 d group (P < 0.05). These results indicate that although juveniles could survive starvation and quickly resume feeding after food became available, the inhibitory effects of extended starvation on growth and molting were not fully compensated during the subsequent 70 d re-feeding period. In conclusion, already feeding 2nd-instar juveniles of T. tridentatus showed strong tolerance to short-term starvation in terms of survival and immediate feeding recovery, whereas prolonged starvation caused significant sublethal effects on subsequent growth and molting. Starvation for ≤6 d had relatively minor effects, while starvation for >6 d resulted in persistent inhibition of growth and developmental progression. These findings suggest that under food-limited conditions, juveniles may adopt a survival-priority strategy by suppressing energetically costly processes such as growth and molting to maintain basic survival. The present study provides a scientific basis for feeding management during artificial seed production of horseshoe crab juveniles and offers important guidance for evaluating and improving the effectiveness of stock enhancement and release programs for T. tridentatus.