Early larval stages of mud crab Scylla serrata were exposed to different concentrations of nitrite (40, 80 and 160 mg L⁻¹ and a control, without added nitrite) and three salinity levels (25, 30 and 35 g L⁻¹) using a static renewal method. No interactive effect of nitrite and salinity was detected. Estimated LT₅₀ in 96-h toxicity tests decreased in all stages with increasing nitrite concentrations in all salinity levels. The 96-h LC₅₀ values of nitrite-N were 41.58, 63.04, 25.54, 29.98 and 69.93 mg L⁻¹ for zoea 1, 2, 3, 4 and 5 respectively. As the larvae grew, they showed a progressive increase in tolerance to nitrite. The toxicity of nitrite to larvae increased with exposure time. The median lethal concentration was not affected by salinity. The chloride component of salinity within 25–35 g L⁻¹ did not seem to be as effective in alleviating toxicity as has been reported in other crustacean species. Based on 96-h LC₅₀ and an application factor of 0.1, the ‘safe level’ of rearing mud crab larvae was calculated to be 4.16, 6.30, 2.55, 2.99 and 6.99 mg L⁻¹ nitrite-N for zoea 1, 2, 3, 4 and 5 respectively.