Empathy is an important psychological capacity that involves the ability to recognize and share emotions with others. In humans, empathy for others is facilitated by having had a similar prior experience. It increases with the intensity of distress that observers believe is occurring to others, and is associated with acute emotional responses to witnessing others’ distress. We sought to develop a relatively simple and fast mouse model of human empathy that resembled these characteristics. We modeled empathy by measuring the freezing of observer mice to observing the footshock of a subject mouse. Observer mice froze to subject footshocks only when they had a similar shock experience 24 hours earlier. Moreover, this freezing increased with the number of footshocks given to the subject and it was accentuated within seconds after footshock delivery. Freezing was not seen in naïve observers or in experienced observers that observed a subject who was spared footshock. Observers did not freeze to a subject’s footshock when they had experienced a swim stress 24 hours prior, demonstrating a specific effect for shared experience, as opposed to a generalized stressor in eliciting observer mouse freezing. We propose that this two-day experimental protocol resembles many aspects of human empathy in a mouse model that is amenable to transgenic analysis of neural substrates for empathy and its impairment in certain clinical disorders.