Anxiolytic drugs exert their pharmacological actions by binding to molecular targets, such as benzodiazepine receptors or 5‐hydroxytryptamine (5‐HT) receptors. Specific genes encode these receptors, or the subunits of which they are formed. Therefore, genetic factors may influence strongly the ability of anti‐anxiety agents to produce their behavioural effects. The literature on this subject is reviewed here, with emphasis on data derived from studies with rodents. We present in a critical way the animal models used in the studies aimed at investigating the genetic basis of the action of anxiolytic compounds, including inbred mice, selected lines, linkage strains or mice generated by targeted mutation. Data show that increased anxiety‐like behaviour is not a predictive factor for increased sensitivity to anxiolytic treatment, and it is possible that gene deletion might not be isomorphic to pharmacological antagonism. It is suggested that the strain differences in anxiety‐like behaviour may be used as a tool in assaying anxiolytic activity of new drugs.