The Fairy Census: Leprechaun Material
The Fairy Investigation Society's two census volumes — the Fairy Census 2014–2017 and the Fairy Census 2017–2023, both compiled by Simon Young and both freely available — contain roughly one thousand first-hand accounts of fairy experiences, collected by questionnaire and published anonymously under numbered entries. They are the largest body of modern material in print, and the archive has reviewed both volumes for leprechaun content.
Nine entries mention the word. Five in the first volume (§96, §218, §319, §393, §448) and four in the second (§659C, §661A, §676, §970). Neither volume's introduction discusses leprechauns, and no entry comes from Ireland.
Census entries are not archive case files. They are the census's material, quoted briefly here under its own numbering for study, and they are not given archive references. The readings below are provisional and follow the classification policy where its terms apply. Readers should consult the entries in full at the source before citing them.
The entries
| Entry | Place and decade | Use of the word | Reading |
|---|---|---|---|
| §96 | England (North East), 1980s | Comparison ("looked a little like a leprechaun") | Two witnesses, independently and in sequence; a three-foot bearded figure in a black hooded cloak, dragging a sack. The dress is not the tradition's. The second witness reported prior sightings of the same kind near his home. MULTIPLE WITNESS / COMPARISON ONLY |
| §218 | US (Arkansas), 1960s | Tentative identification ("what I guess you could call a Leprechaun") | A two-foot elderly figure in browns and greens with a sack over the shoulder, which departed down a hole beneath a peach tree; the witness, then a child, states the hole had not been there before. The witness's grandmother advised leaving the little people alone, which the archive records as the most traditional element in either volume. SINGLE WITNESS / CHILDHOOD / SITE FEATURE REPORTED |
| §319 | US (New England), 1980s | Assumption ("I just always assumed it was a leprechaun") | A hand-sized figure observed running along the bottom of a swimming pool; gone on second look. The catalogue's Greater Web-Footed type rests on a single 1987 letter, and this account was not treated as corroboration. SINGLE WITNESS / SUBMERGED / GONE ON SECOND LOOK |
| §393 | US (Virginia), 1990s | Identification among alternatives ("a sidh or leprechaun") | A three-foot figure in a night road, avoided by swerving. The witness reports prior familiarity with the folk traditions, which assists the vocabulary and complicates the identification in equal measure. SINGLE WITNESS / ROADSIDE / PRIOR FAMILIARITY |
| §448 | Sweden (near Stockholm), 2000s | Hypothetical translation ("in Ireland, he might have been called") | A small man in a workman's apron, carrying a satchel, crossing a field on evident business; gone at the field edge. The witness reports no fear and no excitement. The account is offered under Sweden's own little-people terms, and the archive receives it that way. SINGLE WITNESS / DAYLIGHT / OCCUPIED |
| §659C | Wales (Powys) | Identification, unqualified ("I saw the leprechaun") | The only entry in either volume that identifies without hedging: a six-inch figure, red-haired, in worn forest-green clothing with waistcoat and hat, observed twice in one night on a bedroom shelf; it danced, and signed for silence. The witness's remark that the green was "not bright green, more forest green" is closer to the older record than to the gift trade; see Dress. SINGLE WITNESS / DOMESTIC / REPEAT OBSERVATION |
| §661A | Wales (Swansea), 2000s | Comparison ("a fairy-looking leprechaun-type thing") | Reported speech — "I know you seen me then" — which is rare in the modern material. The account is third-hand, which is not. Both facts are recorded. THIRD-HAND / SPEECH REPORTED |
| §676 | US (Arizona), 1970s | Comparison in dress only ("more like a leprechaun than a fairy") | An eight-to-ten-inch elderly figure observed at close range on the witness's own blanket; grey, soft pointed hat, clothing read as eighteenth-century or older. The comparison concerns costume, not kind. SINGLE WITNESS / DOMESTIC / CLOSE RANGE |
| §970 | Germany, 1998 | Resemblance ("he resembled most closely a leprechaun") | A figure of about one metre seen in an office chair for a few seconds; the witness matched the appearance to pictures found afterwards, and states she had not read about leprechauns beforehand. Her colleague had lately returned from a holiday in Ireland. The archive records this because the census does. SINGLE WITNESS / INDOORS / RESEMBLANCE MATCHED AFTERWARDS |
What the material shows
Of the nine, one identifies without qualification (§659C). The rest compare, assume, translate, or qualify: "what I guess you could call," "sort of like," "a sidh or leprechaun," "might have been called." In the modern material the word is used far more often as a familiar point of reference than as a considered identification, and the archive reads the census accordingly. Three of the nine are American, two Welsh, one English, one Swedish, one German; none is Irish.
This supports the position stated on the sightings page: leprechaun-specific reports are a small minority even within collections devoted to fairy experiences, and a sparse archive is what the evidence looks like.
Editorial notes
Three accounts, from three countries and across five decades (§96, §218, §448), independently describe the figure carrying a sack, bag, or satchel over the shoulder. The detail is not prompted by the census questionnaire, is not remarked on by the census's editor, and is not part of the popular image, which prefers cobbler's tools and a purse. The archive has opened a note on it, in the same spirit as the directed-attention material in INT/0001, INT/0002 and INT/0005: recorded, cross-referenced, and left.
Four accounts (§96, §319, §659C, §970) involve the figure being absent when looked at a second time, or departing within seconds of being observed. This is consistent with the visibility problem the archive has noted in its own case files, and it is added to that note without further comment.
The archive is grateful to the census's compiler and to its witnesses. Inclusion in this review records that an account was published; it does not constitute endorsement of any interpretation, the witnesses' or ours.