In this paper, an accurate method to evaluate the prompt neutron reproduction time for a reflector dominated fast critical reactor, ELECTRA, is discussed. To adequately handle the problem, explicit time dependent Monte Carlo calculations with MCNP, applying repeated time cut-off technique, is used and compared against the σ ∼ 1/v time dependent absorber method, applying artificial cross section data in the Monte Carlo code SERPENT. The results show that when a reflector plays a major role in criticality for fast neutron reactor, the two methods predict different physical parameters (Λ = 69 ± 2 ns and Λ = 83 ± 1 ns for time cut-off and the 1/v method respectively). The reason is explained by applying Avery-Cohn's two-region prompt neutron model.