Practice with national-level exam (FACT, FACT Plus, NET, CUET, etc.) mocks, learn from structured notes, and get your doubts solved in one place.
The screening chemistry every drug analyst runs before any instrument: Marquis (opioids and amphetamines), Mecke (opioids), Duquenois-Levine (cannabinoids), Mandelin (amphetamines), cobalt thiocyanate (Scott test for cocaine), and the false-positive profiles, NIK kit limitations and roadside-screening implications that decide whether a presumptive result can be quoted in court.
Last updated:
A presumptive colour test does not identify a drug. That sentence needs to sit with you before anything else, because the most consequential misuse of a Marquis reagent result in modern forensic chemistry is the misimpression, carried from courtroom to courtroom in many jurisdictions, that a colour change is the same as an identification. It is not. A colour test indicates that a substance belongs to a broad chemical class. It provides probable cause for further investigation. Done well, it narrows a field analyst's decision tree and protects an evidence exhibit from being consumed in confirmatory analysis before a preliminary direction is established.
The practical value of these reagents is enormous. At a seized-drug scene, when the forensic chemist or police officer in the field has a white powder, a green vegetable matter, or a pressed tablet and needs to decide within minutes whether the exhibit warrants custody, transport, and laboratory allocation, a presumptive colour test provides that decision support in seconds, without instruments, for pennies per test.
The chemistry underlying each major test has been known for over a century. Edmund Marquis described his formaldehyde-in-sulphuric-acid reagent in 1896. The Duquenois test for cannabis dates to Pierre Duquenois's work in 1937, later modified by Henry Levine at the New York City Department of Hospitals in 1950. These are not new tools. What is new, and what has driven the most significant legal and scientific re-evaluation of their use since 2010, is the systematic documentation of their false-positive profiles, the case-law in multiple jurisdictions on what a field officer can and cannot assert from a colour test result, and the publication of SWGDRUG's classification of colour tests as Category C (lowest analytical specificity) in the framework that now governs drug identification in most accredited laboratories worldwide.
This topic covers the chemistry of the six major presumptive colour tests, the reagents that generate each colour, the drug classes they reliably respond to and the drug classes and common adulterants that produce the same colour falsely, the three-step Scott cobalt thiocyanate protocol for cocaine, the NIK polytesting kit system and its limitations, and the legal landscape across the US, UK, India and the EU on whether a field officer can characterise a seized substance as a controlled drug on the basis of a colour test alone.
Test yourself on Forensic Chemistry with free, timed mocks.
Practice Forensic Chemistry questionsThe most widely used presumptive drug test in the world is a 130-year-old reaction between a primary amine or aromatic ring and an acidified aldehyde, and it still leads to wrongful arrests when investigators treat its colour as an identification.
The Marquis reagent is a solution of formaldehyde (HCHO) in concentrated sulphuric acid (H2SO4), typically prepared at a 1:25 to 1:50 v/v ratio. The reaction mechanism is an electrophilic aromatic substitution: the formaldehyde, activated by the strongly acidic medium, attacks the aromatic ring of an indole, phenethylamine, or opioid alkaloid structure, forming a coloured condensation product through a series of oxidation and cyclisation steps. The acid also protonates basic nitrogen groups, which drives secondary colour-forming reactions in amphetamine-class compounds.
The colour response pattern for opioids is well established across multiple peer-reviewed validations. Morphine and codeine produce an intense orange-to-red colour that rapidly progresses to dark purple. Heroin (diacetylmorphine) gives an orange-to-purple transition that is slightly slower than morphine because the acetyl groups must hydrolyse in the acid medium before the core morphine skeleton reacts. Methamphetamine produces an orange colour. MDMA (3,4-methylenedioxymethamphetamine) gives a distinctive orange-to-brown-to-black sequence that experienced analysts describe as a "burnt orange developing into an almost black final colour" at two to three minutes.
Amphetamine itself gives a red-orange colour, slower and less intense than MDMA, which provides a rough in-field distinction between the two, though it is not reliable enough for identification. PMA (para-methoxyamphetamine), a phenethylamine frequently substituted for MDMA in pressed tablets, gives a very similar orange-to-dark-brown Marquis result and cannot be distinguished from MDMA by Marquis alone. This substitution is clinically relevant because PMA has a much narrower therapeutic margin than MDMA and has been directly implicated in multiple deaths from hyperthermia in the UK, Australia and Germany when users consumed PMA-containing tablets under the assumption they were MDMA.
The false-positive problem with Marquis is significant. Diphenhydramine, the antihistamine present in over-the-counter cough medicines and sleep aids (marketed as Benadryl in the US, UK and India), produces an orange colour with Marquis that an inexperienced analyst can mistake for an amphetamine response. Procaine and lidocaine, both common cocaine cutting agents, produce orange-to-dark responses on Marquis. DXM (dextromethorphan), present in cough syrups at concentrations of 15-30 mg per dose, produces a reddish-brown Marquis colour that overlaps with opioid responses. A comprehensive 2012 study by Philipp Vorce and colleagues published in the Journal of Analytical Toxicology documented 32 common over-the-counter pharmaceutical compounds that produced colour responses in the primary drug-class colour range for at least one of the six major presumptive reagents.
No single reagent is interpreted in isolation in competent casework; these three reagents form an orthogonal panel that narrows the differential when Marquis alone cannot distinguish.
The Mecke reagent is selenious acid (H2SeO3) dissolved in concentrated sulphuric acid. The selenium in its +4 oxidation state is a more selective electrophile than the Marquis aldehyde-acid combination, making Mecke particularly responsive to the aporphine and morphinan opioid skeletons. Morphine and codeine give a yellow-to-green-to-blue colour sequence with Mecke, progressing over two to four minutes. Heroin gives a blue-green response. MDMA gives a blue-green to black. Amphetamine gives essentially no colour or a faint yellow, which is a useful negative discriminator when a Marquis orange result is ambiguous.
The practical value of combining Marquis and Mecke on a single exhibit is that the pair narrows a phenethylamine from an opioid. A substance that gives orange on Marquis and blue-green on Mecke is more consistent with an opioid skeleton than with a simple amphetamine. Neither result confirms either identity, but the combination substantially limits the class.
The Mandelin reagent is ammonium metavanadate (NH4VO3) in concentrated sulphuric acid. It is particularly useful for detecting amphetamines and ketamine. Amphetamine produces an orange to red-brown. Methamphetamine gives a reddish-orange. MDMA gives a dark blue to black. Cocaine produces a yellow-orange. Heroin gives a brownish-black. The Mandelin reagent's sensitivity to PMA (orange to khaki) is different enough from MDMA (dark blue to black) that the Mandelin-Marquis-Mecke panel has been recommended in several EMCDDA harmonisation documents for in-field distinction of MDMA from PMA-containing tablets, though the distinction requires an experienced observer and a standardised colour reference card.
The Froehde reagent uses molybdic acid (MoO3 dissolved in H2SO4). It gives particularly vivid responses to opioids: morphine yields a violet to blue sequence; codeine gives similar but less intense. Froehde is especially useful for identifying opioids in complex heroin street samples containing multiple adulterants, where Marquis alone may give a muddied colour from mixed alkaloids. The Liebermann reagent (potassium nitrate in H2SO4) is less widely used in routine drug casework but provides a useful supplementary response for cocaine (blue-green) and heroin (brownish-red).
Simon's test is a specific confirmatory presumptive test for secondary amines, used primarily to distinguish methamphetamine (a secondary amine, N-methyl group on nitrogen) from amphetamine (a primary amine). The test uses two reagents: sodium nitroprusside (Na2[Fe(CN)5NO]) in ethanol and sodium carbonate. A secondary amine produces a blue colour that indicates the presence of an N-methyl group characteristic of methamphetamine and MDMA. Primary amines such as amphetamine give no colour reaction. Simon's test was the standard tool used in Australian state police laboratories throughout the 1990s and 2000s for distinguishing amphetamine seizures before GC-MS confirmation, and it remains in use in several national laboratories in Southeast Asia as a presumptive secondary amine screen.
The Duquenois-Levine is the only presumptive drug test that involves a liquid-liquid partition step, and that extra step is precisely what makes it more selective than any single-reagent colour test.
The Duquenois-Levine test for cannabinoids is structurally more complex than the single-reagent colour tests discussed above. It involves three separate chemical steps, and the critical selectivity arises from the third step, a liquid-liquid partition between an aqueous phase and chloroform.
Reagent 1 is the Duquenois reagent: vanillin (4-hydroxy-3-methoxybenzaldehyde) and acetaldehyde dissolved in ethanol. A small amount of the test substance is dissolved in this reagent and the mixture is allowed to stand. For cannabis-containing material (containing cannabinoids including THC, CBD, CBN), a blue-grey to lavender colour develops in this first step.
Reagent 2 is concentrated hydrochloric acid. Addition of HCl to the Duquenois-coloured solution intensifies and often darkens the colour. This step increases the colour density in cannabinoid-positive samples.
Reagent 3 is chloroform. Addition of chloroform and agitation of the biphasic mixture causes a liquid-liquid partition. In a positive Duquenois-Levine result for cannabis, the purple colour partitions into the chloroform (lower) layer, leaving the aqueous (upper) layer colourless or nearly so. This partition step is the key selectivity element: many compounds that produce colour in Reagent 1 alone (various polyphenolic plant materials, some flavonoids) do not transfer their colour to the chloroform layer because the coloured product is insufficiently lipophilic.
The test responds to the whole cannabinoid family: THC, CBD, CBN, THCV, and the synthetic cannabinoids of the classical naphthoylindole class (JWH-018, JWH-073) also give positive Duquenois-Levine results in published testing. Some synthetic cannabinoids of the fluorinated indazole class (5F-MDMB-PINACA, AMB-FUBINACA) have been reported as negative or weakly positive on Duquenois-Levine, a significant limitation when laboratory intake involves presumptive screening of suspected synthetic cannabinoid blends.
False positives are documented but less common than with Marquis. Certain aromatic hydrocarbons, some flavonoid-rich plant extracts, and nutmeg oil (which contains myristicin, a phenylpropanoid) have been reported to produce blue-purple colours in the Duquenois step that partially partition into chloroform. The DEA's own published validation (cited in the DEA Color Test Reagents/Kits document, 2000) notes that the test should not be interpreted as specific for THC or any individual cannabinoid; it identifies the cannabis botanical class with reasonable specificity when the partition step is performed correctly.
In the US, the Duquenois-Levine test was historically accepted by courts in a number of state jurisdictions as presumptive evidence that a material was cannabis, though the 2012 US Supreme Court decision in Melendez-Diaz v. Massachusetts (establishing the Confrontation Clause requirement for analyst testimony) changed the landscape for how any laboratory result, presumptive or confirmatory, is presented. In India, forensic examiners at CFSL Hyderabad and state FSLs use the Duquenois-Levine as part of the initial screening panel for cannabis exhibits under NDPS Act casework, with GC-MS confirmation required before expert evidence is filed.
The Scott test is the only presumptive cocaine protocol that requires three sequential steps for a positive result, and that sequential design is the feature that gives it more specificity than any single-step colour test.
The Scott test (cobalt thiocyanate test) is the standard field presumptive test for cocaine and its salts. The chemistry involves the reaction between cobalt (II) thiocyanate (Co[SCN]2 in aqueous solution) and cocaine, forming the lipophilic complex cobalt(II) thiocyanate-cocaine, which has a distinctive blue colour in the organic phase after extraction.
The three-step Scott test protocol, as standardised in the UNODC ST/NAR/13 manual and adopted by the UK Forensic Science Service (now Forensic Science International), DEA laboratories and CFSL, proceeds as follows.
Step 1: Add the test substance to the cobalt thiocyanate reagent in water. A blue colour forms in the aqueous phase for a broad range of bases and amines, not just cocaine. This first blue is not diagnostic by itself; it simply indicates a basic compound is present.
Step 2: Add dilute hydrochloric acid. The blue colour in the aqueous phase disappears in the presence of acids for non-cocaine compounds, because the cobalt-amine complex is pH-sensitive. For cocaine, the blue persists briefly or changes to a rose-pink.
Step 3: Add chloroform and mix. The blue cobalt-cocaine complex partitions into the chloroform (lower) layer, giving a blue organic phase. The aqueous (upper) layer becomes colourless or pale pink. A positive Scott test requires: blue in Step 1, colour change or persistence through Step 2, and blue chloroform layer in Step 3.
The most clinically important false positive for the Scott test is diphenhydramine (again). Diphenhydramine, procaine, lidocaine, and phenothiazine antipsychotics (such as promethazine, used in India and globally as an antiemetic) can give a blue-in-chloroform result in the Scott test. A 2009 study in the Journal of Forensic Sciences by Phinney and Forrest documented that 9 of 14 common adulterants and cutting agents present in cocaine street samples gave a blue chloroform layer in the Scott test, with diphenhydramine and lidocaine producing the most intense false-positive results.
This false-positive profile has direct legal implications. In the United States, the National Academies of Sciences, Engineering, and Medicine 2016 report "Strengthening Forensic Science in the United States" cited colour tests specifically as examples of methods with unacceptable false-positive rates when used without confirmatory analysis. In England and Wales, Crown Prosecution Service guidance on drug evidence requires that any charge under the Misuse of Drugs Act 1971 be supported by confirmatory analytical evidence (typically GC-MS or LC-MS/MS) from an accredited laboratory; a field officer's Scott test alone does not constitute sufficient evidence for charge. In India, the NDPS Act 1985 and subsequent Supreme Court rulings (including State of Kerala v. Rajasekharan [2006] and Yusuf v. State of Kerala [2011]) have established that chemical analysis from an FSL is required for prosecution; a field test is not a substitute for the FSL report.
A single tube NIK test deployed at a traffic stop does not constitute drug identification, and at least two US exonerations since 2016 have involved misidentifications made with colour tests alone.
The NIK Public Safety product line, manufactured by Safariland Group and distributed widely to law enforcement agencies across North America, Australia and parts of Europe, packages each major presumptive colour test in a sealed ampoule system. An officer crushes the ampoule, mixes the reagent with a small amount of the suspect substance, and reads the colour against a printed reference card. The kits are designed for field use, they require no technical training beyond following the pictorial instructions, and their simplicity is their commercial value.
Their analytical limitations are also well documented. The 2016 report from the American Civil Liberties Union (ACLU), "Unreliable and Biased: The Discriminatory Impact of Drug Field Tests," analysed court records from multiple US jurisdictions and documented numerous instances in which suspects were charged and sometimes convicted on the basis of colour test results that subsequent laboratory analysis revealed to be false positives. Common substances producing false positives in the ACLU dataset included eucalyptus oil, chocolate, donut glaze, and legal dietary supplements, all of which produced a positive Marquis or cobalt thiocyanate test at field conditions.
The polytesting principle, established in SWGDRUG's Recommendations documentation and in the UNODC ST/NAR/13 guidelines, addresses this limitation by specifying that no presumptive identification should be based on a single colour test. The recommended NIK polytesting approach sequences tests so that the results are read cumulatively. For example, for cocaine: Test 8 (cobalt thiocyanate) is run first; a blue colour moves to Test B (Scott modification) for the three-step partition; the combination of the two tests is more selective than either alone. Several US police departments (including Los Angeles County Sheriff's Department in a 2020 protocol revision) have moved to mandatory double-testing protocols after high-profile false positive incidents.
The temperature and concentration dependence of colour test results is a significant analytical confound that does not appear on the NIK colour reference cards. Ambient temperature affects the rate and sometimes the final endpoint of the colour reaction: a Marquis test performed at 5 degrees Celsius (a scenario relevant to a Canadian or UK winter traffic stop) may show a delayed and less intense colour compared to the same test at 25 degrees. The concentration of the test substance in the ampoule depends on the amount the officer adds, which is not controlled. For dilute samples or highly cut street drugs, the colour may be too pale to match the reference card confidently.
The 2018 case of Amy Albritton in Harris County, Texas, USA, achieved broad coverage in ProPublica's investigation "How a Positive Drug Test Can Land You in Jail and Ruin Your Life": Albritton's car was searched, a white substance was found, a field test produced a positive colour response, she pleaded guilty to drug charges to avoid trial, and she was sentenced. Subsequent laboratory testing found no controlled substance. She was exonerated. The case prompted Harris County to change its policy to require laboratory confirmation before any plea offer is accepted in drug cases.
The UK's National Police Chiefs' Council (NPCC) Drug Testing Policy, updated in 2021, explicitly states that Marquis and Scott test results "are not sufficient to support a charge under the Misuse of Drugs Act 1971 without corroborating laboratory analysis." The EMCDDA's Drugs Monitoring Network guidance document EMCDDA Insights 15 (2013) similarly characterises colour tests as providing "indicative, not confirmatory" results and recommends they be combined with at least one Category B or Category A method (in SWGDRUG terms) before a court submission.
The words an analyst uses when reporting a presumptive test result matter precisely as much as the chemistry, because vague language has been used both to overstate what a field test proves and to dismiss laboratory analysis that followed proper protocol.
The legal constraints on reporting presumptive colour test results have been shaped by a sequence of appellate decisions across multiple jurisdictions. The US Supreme Court's Melendez-Diaz v. Massachusetts (2009, 557 U.S. 305) held that forensic laboratory reports are testimonial evidence and that the analyst who performed the test must be available for cross-examination under the Confrontation Clause of the Sixth Amendment. Although Melendez-Diaz concerned laboratory results rather than field tests, its effect on drug casework was to raise the evidentiary standard for all analytical results and to increase judicial scrutiny of the distinction between screening and confirmatory tests.
In England and Wales, the Forensic Science Regulator's Codes of Practice and Conduct (FSR-C-100, Issue 5, 2017) define the requirements for forensic science activities conducted in support of the criminal justice system. Code C7 on drug analysis specifies that identification of a controlled drug requires an analytical method with a level of scientific discrimination consistent with SWGDRUG Category A or B. A colour test result, unaccompanied by any other analytical data, falls below this threshold.
In Australia, the Uniform Evidence Law and its application in Victoria, New South Wales and South Australia courts has generated a body of case law on expert evidence in drug matters. The High Court of Australia in R v. Aporo [2019] (VSCA 133) addressed the admissibility of expert evidence in drug cases and, while the case concerned methamphetamine identification by an expert witness, the court's discussion of the difference between screening and confirmation evidence is instructive: the expert must be able to state the basis of the identification, including its limitations.
In India, the chain of custody provisions under the Bharatiya Nagarik Suraksha Sanhita (BNSS) 2023 (which replaced the CrPC) require that seized articles be forwarded to the FSL for chemical examination. The FSL report, signed by the examiner, constitutes the documentary evidence under Section 293 of the CrPC (now mirrored in BNSS) and is treated as expert evidence. The NDPS Act 1985 itself, in Section 52A, requires that inventory and certification of seized drugs be done under the procedure laid down, and subsequent Supreme Court decisions (Bharat Chaudhary v. State of Bihar, 2003; Noor Aga v. State of Punjab, 2008) have reinforced that the integrity of the FSL analysis is central to the prosecution case. A field colour test result does not substitute for this statutory requirement.
The recommended reporting language for a presumptive test result, per the SWGDRUG Standards and Guidelines (most recently revised in 2019) and per the ENFSI Drugs Working Group best practice manual, is as follows: "Presumptive colour testing of [exhibit reference] with [reagent name] produced a [colour] result consistent with the presence of [drug class]. This result is indicative but not conclusive. Confirmatory analysis using [Category A method] is required before a positive identification can be stated."
This language performs two functions. It accurately represents what the test does (a probability statement about drug class, not an identification). And it creates a record that the analyst understood the test's limitations at the time of the report, which protects the analyst and the laboratory from a defence challenge claiming that a misidentification was made carelessly.
| Reagent | Primary chemistry | Positive drug class | Colour | Key false positives |
|---|---|---|---|---|
| Marquis | HCHO + conc. H2SO4 (electrophilic aromatic substitution) | Opioids, amphetamines, MDMA | Opioids: purple; Amphetamines: orange; MDMA: orange-black | Diphenhydramine (orange), procaine, DXM, lidocaine |
| Mecke | H2SeO3 + conc. H2SO4 (selenium-based electrophile) | Opioids, MDMA, cathinones | Opioids: blue-green; MDMA: blue-black; Amphetamine: faint yellow | Some cathinones, methcathinone |
| Mandelin | NH4VO3 + conc. H2SO4 (vanadate oxidation) | Amphetamines, cocaine, heroin, ketamine | Amphetamine: orange-red; MDMA: dark blue; Cocaine: yellow-orange | PMA (orange-khaki, mimics amphetamine) |
| Simon's | Na2[Fe(CN)5NO] + Na2CO3 (secondary amine detection) | Secondary amines: methamphetamine, MDMA | Blue (secondary amine positive); No colour (primary amine) | Other secondary amines including MDEA |
| Duquenois-Levine | Vanillin + acetaldehyde + EtOH, then HCl, then CHCl3 partition | Cannabinoids (THC, CBD, CBN) | Purple in CHCl3 layer | Some flavonoids, nutmeg oil; most classical synth cannabinoids positive |
| Scott (cobalt thiocyanate) | Co(SCN)2 aqueous, then HCl, then CHCl3 partition | Cocaine and its salts | Blue in CHCl3 layer after 3-step protocol | Diphenhydramine, lidocaine, procaine, promethazine |
| Froehde | MoO3 in conc. H2SO4 (molybdenum oxidation) | Opioids (morphine, codeine) | Violet-blue (morphine); orange-brown (codeine) | Some plant alkaloids |
| Liebermann | KNO3 + conc. H2SO4 (nitration) | Cocaine, some opioids | Cocaine: blue-green; Heroin: brownish-red | Various polycyclic aromatic compounds |
A white powder produces an orange-to-black colour sequence with Marquis reagent and a blue colour with Simon's test. Which interpretation is most consistent with these results?