The aversive aspect of pain constitutes a major burden faced by pain patients. This has been recognized by the pain research community, leading to the development of novel methods focusing on affective-motivational behaviour in pain model animals. The most common tests used to assess pain aversion in animals require cognitive processes, such as associative learning, complicating the interpretation of results. To overcome this issue, studies in recent years have utilized unconditioned escape as a measure of aversion. However, the vast majority of these studies quantify jumping – a common escape behaviour in mice, but not in adult rats, thus limiting its use. Here, we present the "Heat Escape Threshold" (HET) paradigm for assessing heat aversion in rats. We demonstrate that this method can robustly and reproducibly detect the localized effects of an inflammatory pain model (intraplantar carrageenan) in male and female Sprague-Dawley rats. In males, a temperature that evoked unconditioned escape following carrageenan treatment also induced real-time place avoidance (RTPA). Systemic morphine more potently alleviated carrageenan-induced heat aversion (as measured by the HET and RTPA methods), as compared to reflexive responses to heat (as measured by the Hargreaves test), supporting previous findings. Next, we examined how blocking of excitatory transmission to the lateral parabrachial nucleus (LPBN), a key node in the ascending pain system, affects pain behaviour. Using the HET and Hargreaves tests, we show that intra-LPBN application of glutamate antagonists reverses the effects of carrageenan on both affective and reflexive pain behaviour, respectively. Finally, we employed the HET paradigm in a generalized opioid-withdrawal pain model. Withdrawal from a brief systemic administration of remifentanil resulted in a long-lasting and robust increase in heat aversion, but no change in reflexive responses to heat. Taken together, these data demonstrate the utility of the HET paradigm as a novel tool in preclinical pain research.