Illicit drug use contributes to substantial morbidity and mortality. Drug scheduling, a legal measure in drug enforcement, is often structured as a hierarchy based on addiction tendency, abuse trends, and harm, but may lack data-driven evidence when classifying substances. Our study aims to measure addiction tendency and use trends based on real-world data. We used the open access database of National Police Agency, Ministry of the Interior in Taiwan and analyzed all daily criminal cases of illicit drugs from 2013 to 2017 and monthly illicit drug enforcement data from the same database from 2002 to 2017. We hypothesized that repeat and frequent use despite legal consequence may be a reflection of addictive behavior, and empirical mode decomposition was applied in analysis to calculate addiction tendency indices and intrinsic 15-year use trends. Our analysis showed heroin has the highest addiction index, followed by methamphetamine. 3,4-Methyl enedioxy methamphetamine, marijuana, and ketamine had lower addictive propensities. This result is consistent with most drug scheduling hierarchies. 15-year use trends of substances were consistent with previous epidemiological studies.