Tobacco products are widely abused in humans, and it is assumed that nicotine is the key substrate in these products that produces addiction. Based on this assumption, several pre-clinical studies have utilized animal models to measure various aspects of nicotine addiction. Most of this work has focused on behavioral measures of nicotine and how other variables contribute to these effects. Here we discuss the most commonly used animal models including, self-administration (SA), place conditioning (PC), and the intracranial self-stimulation (ICSS) paradigms in rodents. The strengths, limitations and procedural variables of these models are reviewed, followed by a discussion of how the animal models have been used to study factors such as age, sex, stress, and the effects of tobacco products other than nicotine. These factors are discussed in light of their influences on human tobacco abuse. The rodent models are evaluated in the context of face, predictive, and construct validity, and we propose that inclusion of factors such as age, sex, stress and other constituents of tobacco aside from nicotine can increase the utility of these animal models by more closely mimicking human tobacco abuse.