This paper inspects changes in crime trends brought about by the COVID-19 pandemic in Tamil Nadu, India. Using Bayesian structural time-series models, the authors study the changes in two metrics of criminal activity – crime registration and distress calls. In addition to absolute changes, the authors analyse the relative changes in one metric with respect to the other during two stay-at-home orders across six categories of crimes: property offences; offences against the human body; cases of missing persons and unidentified dead bodies; vehicle- and traffic-related offences; crimes against women, children and elders; and emerging crimes. While there was an overall decrease in most types of crime, results show that each category was impacted by restrictions in human mobility in different ways; emerging crimes, in particular, increased significantly during both periods. The crime and call trends exhibited a synchronous pattern with respect to a majority of offences except in the category of crimes against women and missing persons, where a contrary trend of decreasing crime registrations and increasing distress calls was noticed during the partial lockdown period. This finding highlights inherent inadequacies such as under-reporting of crimes by victims and non-registration of crime by the police. Understanding these inadequacies could potentially enhance preparedness leading to improved public service delivery by reordering priorities. Breaking down these trends at a crime-specific level assists law enforcement practitioners in gaining insights on crime motivators and enablers, thus contributing towards more effective crime prevention under ordinary circumstances.