The case file

On 21 April 1988 a single strip was cut from one corner of the cloth, next to the spot sampled in 1973, away from patches and burn marks. Arizona, Oxford and Zurich each got a piece of about 50 mg, plus three control cloths of known age. The lab means were 646 ± 31 BP (Arizona), 750 ± 30 BP (Oxford) and 676 ± 24 BP (Zurich), giving an unweighted mean of 691 ± 31 BP. Calibrated, that is AD 1273-1288 at 68% and AD 1262-1312 or 1353-1384 at 95%, which Nature rounded to AD 1260-1390. All three controls came out close to their known ages. The paper, published in Nature in February 1989, called the result conclusive evidence that the linen is medieval.

Why this rating: Peer-reviewed, three labs, working controls, and no tested contamination mechanism has been shown to shift the date by 13 centuries.

Nature itself reported a chi-squared of 6.4 on 2 degrees of freedom for the shroud, significant at the 5% level, while the controls passed easily. Riani et al. (2013) found significant differences in lab means (p = 0.0400 unweighted, 0.0043 with original weights) and an age trend along the strip across all 387,072 possible subsample layouts. Casabianca et al. (2019) used raw lab data released by the British Museum in 2017 and found a Ward and Wilson test value of 15.99 against a critical value of 5.99, and an OxCal agreement of only 41.8% on the published dates. Di Lazzaro et al. (2020) reported a spatial gradient of roughly 50 radiocarbon years per cm. Walsh and Schwalbe (2020) found that shifting two labs by about 10 years restores homogeneity, and 88 years brings all results into agreement.

Why this rating: The statistical problem is real and peer-reviewed, but it widens the uncertainty rather than pointing to the first century.

Rogers, a STURP chemist, published in Thermochimica Acta 425 (2005). He reported a yellow-brown gum coating with dye lakes and cotton fibres in threads said to come from the radiocarbon sample, and concluded 'the worst possible sample for carbon dating was taken.' His threads came from Luigi Gonella in 2003, who was not authorised to remove shroud material, so their provenance is uncertain. Freer-Waters and Jull (2010) examined an actual leftover Arizona fragment and found only a few cotton fibres, no dye and no coating. A further Arizona fragment, kept by the late Douglas Donahue, was examined in a 2026 Heritage Science paper, which press reports say found no evidence of a repair or foreign material.

Why this rating: Peer-reviewed, but rests on threads of doubtful provenance and is contradicted by direct study of documented fragments in 2010 and 2026.

Several mechanisms have been proposed since 1988. A review in Textiles (2025) notes that roughly 40% or more modern carbon would be needed to turn a first-century cloth into a 14th-century date, and that a proposed fire carboxylation effect was ruled out by three later studies. Oxford's Christopher Ramsey tested the carbon monoxide idea in 2008 and found no significant reaction, with sensitivity to offsets of under a year. A bacterial 'bioplastic' coating was proposed by Garza-Valdes and Mattingly in 1996. A neutron hypothesis linked to an earthquake appeared in a journal paper later withdrawn in 2015 over a compromised editorial process.

Why this rating: Each specific mechanism that has been tested failed, and the amounts required are implausible.

The team used a single thread about 0.5 mm by 1 mm from near the 1988 sampling corner. Its ageing parameter was 0.60 ± 0.02, which they say cannot be reached by seven centuries at Turin conditions (12.5 °C, 75% RH, giving 0.027 or less). Matching it needs about 20.2 ± 0.8 centuries of ageing if the cloth spent 13 unrecorded centuries at 22.5 °C and 55% RH, or at about 20 °C and 75% RH. If the average was above 23 °C, the age falls under 20 centuries. The authors call the result 'very probable' and say more samples are mandatory.

Why this rating: Single sample, unvalidated method, and the result is driven by assumed storage conditions.

Results at 95% confidence were 300 BC ± 400 years by FT-IR (752 BC before a fire correction), 200 BC ± 500 by Raman, and AD 400 ± 400 by mechanical tests, averaged to about 33 BC ± 250. A 2015 Monte Carlo reanalysis of the mechanical data gave AD 279 ± 216. Calibration used nine ancient textiles plus modern controls. The fibres came from dust vacuumed from the back of the cloth by Giovanni Riggi.

Why this rating: Uncertain fibre provenance, unvalidated calibration and very wide error bars.

Vanillin is a breakdown product of lignin in flax that slowly disappears. Rogers reported 0% retained vanillin in the main cloth versus about 37% in the threads he attributed to the radiocarbon area. Using estimated loss rates, he put the cloth at 1,300 to 3,000 years old.

Why this rating: Untested kinetics and a known heat event make the estimate unreliable.

STURP's 1981 summary states that no pigments, paints, dyes or stains were found on the image fibrils, and that X-ray fluorescence and microchemistry rule out paint. The colour sits on the outermost two or three fibres of a thread; the colour itself is fairly uniform and darkness depends on how many fibres are coloured per area. Rogers and Arnoldi describe the coloured layer as too thin to resolve with a light microscope. A 2025 peer-reviewed review still calls image formation 'a mystery.'

Why this rating: Well documented by direct study in 1978 and confirmed by later reviews; what it implies for age is a separate question.

McCrone examined fibres from 32 tape samples taken in 1978. He reported red ochre on 20 image and blood tapes, vermilion only on 11 blood tapes, and neither on 12 non-image tapes. He said forensic blood tests on blood-image fibres were negative. He published in The Microscope (1980, 1981) and Accounts of Chemical Research (1990) and dated the painting to about 1355.

Why this rating: A serious expert claim, but contradicted by several independent chemical and physical analyses of the same 1978 material.

Pia took the photographs on the evening of 28 May 1898 using electric lamps. The negative revealed facial detail hard to see by eye. Some accused him of tampering until Giuseppe Enrie's 1931 professional photographs confirmed the effect.

Why this rating: A real and striking property, but reproducible in rough form by simple contact methods.

In 1976 a VP-8 Image Analyzer rendered an Enrie photograph as a coherent 3D form, while ordinary photos give a distorted jumble. In Applied Optics 23 (1984), the correlation from a 13-point sample was r² = 0.60, with a 95% interval of 0.13 to 0.86. The image fades out at about 3.7 cm of cloth-body distance. An ordinary photo of a plaster face gave r² = 0.17, while a distance-coded test image gave 0.98. The authors judged that contact alone, diffusion, simple radiation and simple bas-relief all fail to fit everything.

Why this rating: Genuine, peer-reviewed property, but the statistics are thin and low reliefs can mimic it.

Cicero Moraes published in Archaeometry on 28 July 2025. He draped virtual cloth over a full 3D body and over a low relief and compared both with the 1931 photographs. The 3D body produced the widened 'Agamemnon mask effect'; the low relief matched. He called the Shroud a masterpiece of Christian art and did not study its materials.

Why this rating: The geometric observation is valid and long known, but the study ignores chemistry and does not show a relief can produce the actual fibre-level image.

Nickell moulded wet cloth over a low relief and rubbed on powdered pigment, reporting about 30 points of similarity. Garlaschelli (2009, published 2010) rubbed an acidic pigment slurry over a volunteer's body and a bas-relief for the face, baked the cloth at about 140-145 °C for 3 hours, then washed out the pigment. He reported a superficial, front-only, negative image that does not fluoresce under UV, while burn marks do.

Why this rating: Shows several features are reachable by art, but no replica reproduces all of them together.

Di Lazzaro and colleagues used a 193 nm laser with 12 nanosecond pulses. Shroud-like yellowing appeared only in a narrow band of about 2 to 4 billion watts per square cm; more ablated the linen. Average colour depth across yarns was about 7 to 26 micrometres, with one fibre showing colour only in a primary wall about 0.2 micrometres thick. Scaling the threshold to a body-sized area gives about 34 thousand billion watts, which the authors say no existing UV source can deliver.

Why this rating: Good lab physics but no evidence such a source existed; the wattage number must not be used as a 'proof'.

Linen washed in Saponaria and left with starch residues would carry reducing sugars to the surface as it dried. Ammonia and amines such as cadaverine and putrescine could react with them at moderate temperature. In tests, ammonia vapour on treated linen gave golden surface colour with clear fibre cores. The authors state this does not prove how the image formed or that the cloth is authentic.

Why this rating: Chemically plausible but never shown to produce an image and lacking supporting residue evidence.

The analysis used computer processing of photographs taken in 2002, when the backing cloth was off. Published in Journal of Optics A 6 (2004). The authors argue the back image is superficial, not soak-through.

Why this rating: Single indirect study, unconfirmed by direct observers or independent reanalysis.

Heller and Adler (1980, 1981) ran about 12 lines of evidence on the 1978 tapes: Soret absorption near 400-450 nm, conversion of heme to fluorescent porphyrin, hemochromogen and cyanmethemoglobin colours, protein tests, albumin tests and enzymes that dissolved the blood coating but left image fibres alone. Image and plain fibres tested negative for protein above the nanogram level. STURP's 1981 summary says the stains are hemoglobin and test positive for serum albumin.

Why this rating: Multiple consistent chemical tests, but old methods, not repeated with modern techniques, and not diagnostic of age.

Heller and Adler used proteases that removed blood coatings within 30 minutes. Rogers and Arnoldi also report image colour absent beneath enzymatically removed stains. A 2026 experimental study by Kelly Kearse found that blood applied to acid-treated linen prevents later coloration in that spot.

Why this rating: Based on limited samples, and new experiments show it does not uniquely fix the order of events.

Forward typing found A and B antigens but not O on bloodstained fibres. Reverse typing found no anti-A or anti-B antibodies. Rh was never tested, so 'AB positive' has no basis. Garza-Valdes later reported only B antigens.

Why this rating: Unpublished in peer review and vulnerable to bacterial false positives; even sympathetic specialists reject the inference.

Heller and Adler (1981) reported azobilirubin colour on some blood particles and a spectral peak near 450 nm. Adler said bilirubin mixed with methemoglobin would look red.

Why this rating: Never quantified or experimentally confirmed.

Using a living volunteer and modern bloodstain pattern methods, they found the hand rivulets need arms at about 45° while forearm flows need nearly vertical arms. The chest side-wound flow looked realistic for a standing man, but the 'belt' of blood across the lower back was judged 'totally unrealistic' for a body lying down.

Why this rating: Peer-reviewed forensic work, but a single-volunteer setup that cannot capture movement or post-mortem flow.

Medical examiner Robert Bucklin described at least 100 bifid scourge marks from shoulders to calves, a ring of scalp punctures, a swollen right cheek, shoulder abrasions, a wrist wound with two diverging rivulets, and a right-chest wound with blood and clear fluid. Frederick Zugibe ran suspension experiments on volunteers and concluded death came from traumatic and hypovolemic shock. STURP concluded in 1981 that the image is of a real human form of a scourged, crucified man.

Why this rating: Consistent expert medical readings, but by sympathetic experts who disagree on details, and not diagnostic of date.

Barbet argued the nail passed through Destot's space in the wrist and severed the median nerve, drawing the thumb into the palm. Zugibe disputed this: he placed entry in the upper palm along the thenar crease with exit at the wrist, and said the thumbs are simply hidden in their resting position. Bevilacqua and D'Arienzo (2015) again place the nail at Destot's space.

Why this rating: The observation is clear; its interpretation is contested among sympathetic experts.

Bucklin counted at least 100 bifid marks from shoulders to calves, absent from the forearms. Zugibe noted that 40 lashes with a three-thonged whip would leave about 120 marks. He read them as small skin breaks with blood, not bruises.

Why this rating: Clear, consistent pattern, but compatible with an informed artist.

Gregory S. Paul (2010) called the figure anatomically implausible, 'exceptionally tall' with a 'shrunken and low browed' head. Nickell pointed to hair falling as if upright and blank zones a true imprint would not have. Rebutters estimate height from the image at roughly 173 to 177 cm, though methods vary.

Why this rating: Measurements from a draped, distorted cloth are too uncertain to support strong conclusions either way.

The cloth is about 4.4 m by 1.1 m, about 0.35 mm thick, with threads about 0.15 mm across. Flury-Lemberg says the 3:1 herringbone marked extraordinary quality in antiquity.

Why this rating: Consistent with antiquity but not diagnostic of any period.

She studied the cloth closely during the 2002 restoration and judged the seam professional. The side strip is the same original fabric. She is one of the most respected textile conservators of her generation.

Why this rating: A single expert parallel, not shown to be unique to antiquity.

Raes, of the Ghent textile institute, studied threads cut on 24 November 1973 near the later radiocarbon corner. The species identification is reported by Ian Wilson. Raes suggested the loom had previously woven cotton.

Why this rating: Compatible with many times and places.

Experimental weaver Antoinette Merete Olsen tested the question in EXARC Journal (2020/4) after a query from researcher Hugh Farey. Wikipedia summarises her result as suggesting a treadle loom was required.

Why this rating: A single experimental study whose full conclusions are not widely discussed; the full paper is not freely available.

Frei sampled in 1973 and 1978 and claimed about 57 to 58 plant types, often to species level. He died before publishing full data. Botanist Avinoam Danin later linked the plant set to the Jerusalem area in spring.

Why this rating: Species identifications were never verified and key ones were disputed by a professional palynologist.

The images were found in enhanced photographs and compared with drawings in Flora Palaestina. Danin mapped 27 of 28 species to the Jerusalem area.

Why this rating: Subjective, unreplicated reading of enhanced photographs; the site should not use it.

The coin reading rests on photo enhancement of an area a few millimetres across and on letters UCAI, explained by a supposed minting error. Frale announced her reconstructed death certificate through the press before peer review. Both claims depend on enhanced photographs rather than direct examination of the cloth.

Why this rating: Readings below the resolution of the weave, an unattested burial custom, and rejection by leading pro-authenticity researchers; the site should not use either claim.

The match used optical properties (refractive index 1.680) and an ion-microprobe comparison by Levi-Setti on a few particles from one tape. The author called it 'a strong suggestion,' not proof.

Why this rating: Tiny sample and limited method; the authors themselves call it suggestive.

Barcaccia et al. (2015) found traces of 19 plant taxa and diverse human mitochondrial lineages and said a medieval European origin was not excluded. Their 2026 Scientific Reports study of the official 1978 samples found the dominant human DNA in three samples matched Pierluigi Baima Bollone, who collected them. It also found wheat, maize, peanuts, bananas, farm animals and red coral, and radiocarbon dates on two threads consistent with the 1534 and 1694 repairs.

Why this rating: Dominated by modern contamination; tells us about handling, not origin.

The Sudarium has no image, only stains. A 2017 UCAM press release by forensic researcher Alfonso Sánchez Hermosilla and colleagues said the two cloths covered the same body with a post-mortem chest wound. The earliest explicit record is the opening of its chest on 14 March 1075.

Why this rating: Own radiocarbon dates are early medieval, history is late and unreliable, and the match claim is not peer reviewed.

The drawing is in the Pray Codex, dated by the funeral sermon to 1192-1195. Daniel Scavone and others argue the circles 'perfectly reproduce' the L-shaped burn holes still visible on the Shroud. If correct, it would place the Shroud, or a careful copy, about 70 years before the earliest radiocarbon limit.

Why this rating: Intriguing parallels, but a plausible non-Shroud reading exists and scholars on both sides disagree.

Eusebius (early 4th century) records Abgar's letter with no image, and Egeria (384) saw letters but no image. Evagrius (about 593) is the first to credit an image with saving Edessa in 544. In 944 the image was moved to Constantinople. Wilson cites a 10th-century text describing a full-body imprint.

Why this rating: Speculative identification rejected by most historians.

De Clari wrote his account after 1204. It is the only Western eyewitness-style description of a cloth with a full figure in Constantinople.

Why this rating: A genuine source but uncorroborated and contested in meaning.

Early pilgrim accounts (4th century) and 6th-century relic lists mention many Holy Land relics but not a figured shroud. The first secure trail starts with Geoffroi de Charny's church at Lirey in 1353 and pilgrim badges with the Charny and Vergy arms. Many other 'true shrouds' circulated in medieval Europe.

Why this rating: Documented gap that fits the radiocarbon date; the alternative routes are speculative.

Pierre d'Arcis wrote to Antipope Clement VII that the Lirey dean had procured a cunningly painted cloth for profit. Clement's bulls of 6 January 1390 allowed showings only if announced as a painting or representation, with no candles, incense or guard.

Why this rating: A real, early hostile source with a papal response, though uncorroborated on the key 'confession'.

Historian Nicolas Sarzeaud published the find in the Journal of Medieval History in August 2025. It is the earliest known written rejection, likely after 1370 and before d'Arcis. Oresme had no personal stake in the dispute.

Why this rating: Independent and early, but brief and without stated evidence.

Freeman (History Today, 2014) cites a 1613 Tempesta engraving and 15th-16th century descriptions showing clearer blood and scourge marks, plus the Lirey badge showing clerics holding the cloth above a tomb. He links it to the 'Quem quaeritis' ritual.

Why this rating: A coherent idea from a serious historian, but it conflicts with the physical study of the image fibres.

Steven Schafersman described STURP as believers with one exception, Walter McCrone. Joe Nickell noted the Guild links. STURP had no experts in medieval art, archaeology or textiles.

Why this rating: Largely an argument about motive; the published data must be judged on their own.

John 19:40 and John 20:5-7 describe linen cloths (othonia) and a separate soudarion. The synoptic Gospels say the body was wrapped in a linen sindon (Mark 15:46).

Why this rating: Depends on translation of a plural Greek noun; does not decide anything.

John Paul II said in Turin on 24 May 1998: 'Since it is not a matter of faith, the Church has no specific competence to pronounce on these questions,' and entrusted the matter to scientists. Benedict XVI and Francis called it an icon. Archbishop Nosiglia said in 2013 there was 'no degree of safety' on Fanti's tested material.

Why this rating: Documented official statements.