Abstract: Emotion has important influence on memory. People usually remember better for emotional stimuli than for neutral stimuli. In previous researches, effects of emotional arousal and valence of targets on memory were discussed in detail. However, so far few researches investigate the influence of the emotional relationship between target and context on memory. Therefore, the purpose of the current research was to examine whether and how affective congruency between target and context during encoding impact on memory retrieval by employing affective priming and learning-recognition paradigms and event-related potential (ERP) measurement. Sixteen right-handed undergraduate students participated in the experiment. The stimuli consisted of 600 ordinary emotional pictures (300 positive and 300 negative pictures) as targets during encoding or retrieval and 240 facial emotional pictures (80 positive faces, 80 negative faces, and 80 neutral faces) as primes during encoding. Among prime pictures, 60 pictures were used twice. The experiment was divided into five blocks, each of which comprised encoding, distracting, and retrieval. In encoding phase, each trial began with a fixation point. Following the fixation, a prime picture was presented for 200 ms, and was replace by black screen for 50 ms. A target picture was then displayed for 1500ms. Participants were asked to judge the pleasantness of the target. Having finished the study phase, participants executed a 60s distracter task. In retrieval phase, each trial also began with a fixation point. A target picture was then presented for 250 ms followed by black screen for 2000 ms. Participants were required to press a button as soon as possible to indicate whether the pictures were presented in previous study phase. EEG was recorded from 64 scalp electrodes using an electrode cap. Behavioral results indicated that accuracy rates of recognition for target pictures that were encoded under congruent priming condition (e.g., positive-positive) were significantly higher than under incongruent priming condition (e.g., negative-positive) and neutral priming condition (e.g., neutral-positive) (ps<0.05). However, affective congruency between prime and target didn’t significantly impact on response times (p>0.05). ERP results from retrieval phase revealed, during 300-500ms after target onset, ERPs evoked by negative targets from congruently priming conditions were significantly more positive than from neutral priming conditions (p<0.05). In addition, ERPs evoked by positive targets from congruently priming conditions were significantly more positive than from incongruently priming conditions (p<0.05) during 550-700ms. These results suggest that the context being consistent affectively with targets improves the processing for targets, resulting in high accuracy rate of recognition and more positive ERP in memory retrieval.