The aim of the current study was to examine the effects of three factors on the emergence of context effects (CEs). The basic assumption was that the way Target and Context words are initially encoded affects their impact on recognition CE. The strength of the memory depends on many factors, including the amount of attention allocated to target and context stimuli when memorizing the information and whether the participants were distracted. The interaction and relationships between these different factors were examined in the present study. First is Congruency between target and context words in terms of the gender of the nouns presented. Second is Attention allocated to the stimuli, whether equal attention so that both are considered targets (T-T) or differential attention allocation so that one is the target and the other is context (T-C). Third is Exposure time of 300 vs. 3000 msec. We hypothesized that CE would be stronger under the T-T vs. T-C attention condition, congruent vs. incongruent learning conditions and short vs. long exposure time. One-hundred and fifteen individuals participated in this study. They were randomly assigned to one of the two Exposure time conditions. Half of the word-pairs were congruent and half were incongruent. Short exposure time in the Congruent T-T condition was associated with CE in terms of hit rates, but not false alarms, with no CE in the incongruent pairs. As predicted, lengthening exposure time reduced CE in terms of hit rates, and congruent relations were associated with greater CE in terms of false alarms, with no influence of encoding type.