Novel psychoactive substances (NPS) are new drugs of abuse. Over the last 10 years 50–100 new NPS have been detected for the first time each year. This has led to numerous deaths and challenges to healthcare providers and law‐makers worldwide. We review preclinical studies of NPS and discuss how these studies have influenced legislative decisions. We focus on the UK legal system but include experiences from Europe. We reviewed manuscripts from 2008–2019 and have summarised the in vitro and in vivo data on NPS, highlighting how these studies define pharmacological mechanisms and how they might predict harm in humans. We found that only a small percentage of NPS have been examined in preclinical studies. Most preclinical studies of NPS focus on basic pharmacological mechanisms (46% of studies reviewed) and/or addictive liability (32%) rather than toxicity and harm (24%). Very few preclinical studies into NPS include data from chronic dosing schedules (9%) or female rodents (4%). We conclude that preclinical studies can predict harm to humans, but most of the predictions are based on basic pharmacology rather than demonstrated toxicity. Some of these studies have been used to make changes to the law in the UK and Europe and perhaps, because of the paucity of toxicology data, most NPS have been placed in the highly dangerous schedule 1 or Class A category in the UK. We suggest that in silico studies and high throughput toxicology screens might be the most economical way forward to rapidly screen the health harms of NPS.