A forensic laboratory can generate excellent analytical data and still fail to identify a substance. The instrument detects a chemical signal, but the compound does not match any entry in the available reference libraries. A small structural modification to a known drug may be enough to create a substance that routine screening methods cannot confidently recognize. This is a growing challenge in the analysis of novel psychoactive substances (NPS), commonly called designer drugs. These compounds can mimic the effects of established drugs while presenting new analytical and regulatory challenges.
In August 2026, the United Nations Office on Drugs and Crime (UNODC) reported 1,525 NPS from 155 countries and territories through its monitoring system. A further 110 substances were reported for the first time during the 2025/26 reporting period, although the most recent figures remain preliminary. For forensic laboratories, the challenge is no longer simply to detect known drugs. It is to investigate substances for which reference standards, toxicological information and validated methods may not yet exist.
Why Conventional Testing Is Not Enough
Traditional drug analysis relies on comparison with known compounds. Analysts examine characteristic signals, retention times and mass spectra, often using certified reference materials to support identification. These approaches remain reliable for established substances when appropriate validated methods are available. Problems arise when an unfamiliar compound enters the sample. Structural changes can alter fragmentation patterns and chromatographic behaviour, preventing a match with existing library entries. Complex mixtures, low analyte concentrations and the presence of adulterants further complicate interpretation. In biological specimens, the parent drug may also decline before collection, leaving metabolites as the principal evidence of exposure. Non-targeted analysis offers a complementary approach. Rather than examining only a predefined list of compounds, analysts investigate a broader range of detectable chemical features. This does not guarantee identification of every unknown, but it increases the likelihood of recognising unexpected substances and generating useful leads.
High-Resolution Mass Spectrometry: Beyond Targeted Screening
High-resolution mass spectrometry (HRMS) has become an important tool for investigating unfamiliar drugs. It measures mass-to-charge ratios with high accuracy, helping analysts determine candidate molecular formulas and interpret fragmentation patterns. Combined with chromatography, it can provide detailed information about substances within complex mixtures. One of its greatest advantages is retrospective analysis. Consider a seized sample examined before its active ingredient is known. Months later, another laboratory identifies a new drug and publishes its characteristic spectral data. If the original HRMS acquisition captured the relevant signals, analysts may be able to search the stored data for evidence of that substance without reanalysing the physical sample.
This capability is particularly valuable because emerging drugs can spread faster than laboratories can develop reference materials and confirmatory methods. HRMS research has increasingly explored retrospective screening, broad-spectrum wastewater analysis and computational approaches to interpreting unfamiliar compounds. However, accurate mass alone rarely establishes a unique molecular structure. Closely related compounds can share molecular formulas and produce similar fragmentation patterns. Depending on the case, additional evidence, an orthogonal analytical technique or an authentic reference standard may be necessary for confirmation.
Additional Dimensions: Ion Mobility and Artificial Intelligence
Ion mobility spectrometry can complement HRMS by separating gas-phase ions according to their movement through a gas under an electric field. This behaviour depends partly on an ion’s size and shape. Certain isomers compounds with the same molecular formula but different structures can therefore exhibit different mobility characteristics. Collision cross section (CCS), a parameter describing an ion’s interaction with the surrounding gas, provides another measurement for comparison. Although ion mobility can help distinguish some isomers, it cannot resolve every structural ambiguity. CCS should be interpreted alongside other analytical evidence rather than treated as a unique identifier.
Artificial intelligence (AI) offers another way to manage the growing complexity of chemical data. Machine-learning models can suggest structural families, predict candidate fragmentation spectra and prioritise signals for further investigation. Such tools may help analysts narrow the possibilities when conventional library searches return no satisfactory match. Their reliability, however, depends on the quality and diversity of training data. A model trained primarily on familiar compounds may perform poorly on genuinely new chemical classes or data collected using different instruments. Before adoption in forensic casework, these systems require independent validation, testing on unfamiliar compounds and realistic mixtures, and transparent reporting of uncertainty. A computational prediction is not a confirmed identification. The most defensible role for AI is to support expert interpretation, not replace it.
Building a Global Early-Warning Network
Analytical innovation is only part of the solution. When a laboratory identifies a new substance, that information must reach other laboratories quickly enough to influence their testing strategies. Shared spectral libraries and quality-controlled databases can help laboratories compare findings, update suspect lists and develop appropriate methods. Consistent data formats, reporting standards and transparent identification criteria are essential for meaningful international collaboration. Wastewater analysis provides another complementary source of intelligence. HRMS can help researchers investigate chemical residues and metabolites in wastewater to identify patterns that may indicate changing substance use. Such findings can support population-level surveillance without directly identifying individual users. Nevertheless, a detected chemical signal does not automatically establish the prevalence of use, and environmental findings require careful interpretation.
Together, evidence from seized materials, biological specimens and wastewater could strengthen early-warning systems. Their effectiveness will depend on reliable analytical data, appropriate interpretation and timely communication between forensic laboratories and public-health authorities.
What the Next Decade Could Bring
Over the next five to ten years, progress is likely to depend on integrating existing technologies rather than relying on a single breakthrough. In the near term, wider adoption of retrospective HRMS analysis could allow laboratories to extract more value from previously collected data. Machine-learning tools may subsequently become useful decision-support systems as their performance is independently validated. Computational predictions of drug metabolism could also help researchers prioritise potential biomarkers for experimental investigation. A longer-term ambition is a connected network in which an unfamiliar spectrum can be compared against international datasets and used to generate candidate structures, likely metabolites and suggested analytical targets. Such a system could reduce delays between a substance’s first detection and its recognition elsewhere, although expert interpretation and confirmatory testing would remain essential.
Achieving this vision requires investment in instrumentation, reference materials, specialist training and accessible data infrastructure. International cooperation is equally important, particularly for laboratories with limited analytical resources.
From Reaction to Anticipation
Forensic drug analysis has traditionally responded after a new substance enters circulation. A compound appears, investigators identify it, and laboratories develop methods to detect it. HRMS, ion mobility, computational tools and shared databases offer opportunities to shorten this process. The objective is not to predict every new drug before it appears. It is to ensure that unfamiliar substances can be investigated efficiently, analytical uncertainty is communicated honestly, and reliable findings reach the wider scientific community without unnecessary delay. The chemistry of designer drugs will continue to evolve. Forensic science must respond not only with better instruments, but also with stronger validation, skilled analysts and a willingness to share knowledge. These are the foundations of a more anticipatory approach to emerging drug threats.











