The N-methyl-D-aspartate receptor (NMDAR) is thought to be essential for synaptic plasticity and learning. However, recent work indicates that the role of this receptor depends on the prior history of the research subject. For example, animals trained on a hippocampus-dependent learning task are subsequently able to acquire new information in the absence of NMDAR activation. The current experiments were designed to identify the types of experiences that lead to NMDAR-independent learning. Using contextual fear conditioning in mice, we find that NMDAR-independent learning is only observed when (1) animals are trained on the same behavioral task and (2) initial learning is successfully encoded into long-term memory.