Learn · Essay · 12 min read · Updated September 2026

Evidence and feng shui

What has actually been studied, what has not, and what the studies that exist can and cannot say about a tradition that was never designed to be tested. The site’s evidence label, explained at length.

Oriented W 264°

Metal

Fig. 01 Sunlight in a bare room. Most of the evidence that touches feng shui is evidence about light.

Every claim on this site carries a label that says what kind of claim it is: tradition, school, interpretation, design observation, evidence, misconception. This essay is the long version of the fifth label. It sets out what has actually been studied that bears on feng shui, what has not, why the two lists look the way they do, and why the answer to “does it work?” is not a verdict but a sorting.

Three kinds of claim

The tradition makes three kinds of claim, and they cannot be tested the same way. The first kind is about the perceptible environment: light, air, damp, noise, a wall behind the bed, a corridor that is dark at its end. These are claims about rooms that a body can check, and most of them have been studied, by people who had never heard of feng shui and were asking about sleep, recovery or stress. The second kind is about correspondences: this direction is Wood, this number is Water, this shape belongs to Metal. These are not claims about outcomes at all; they are the grammar of the system, and asking whether the south-east is really Wood is like asking whether Tuesday is really the second day of the week. The third kind is about fortune: that a bed in this direction, a door in that sector, a chart of this period will bring health or money or its opposite. These are claims about outcomes, and none of them has been tested, for reasons that are worth setting out.

What has been studied

Fig. 01 The tradition’s claims against the kind of evidence that exists for each. A red dot marks the strongest kind of study that bears on the claim; a ring at the left marks a claim nobody has tested. The studies in the upper rows are not studies of feng shui. They are studies of light, noise, damp, daylight, views, clutter and room shape that happen to test what the tradition also recommends. The four claims at the bottom are the ones the tradition is best known for, and the ones with nothing behind them but the tradition.

The upper rows of the ladder are the tradition’s claims about rooms, and the evidence for them is good because it was not gathered for the tradition’s sake. Room light before bed suppresses melatonin and shortens its duration (Gooley and colleagues, 2011); an evening on a lit screen delays sleep and dulls the next morning (Chang and colleagues, 2015; Cajochen and colleagues, 2011); sleep depends on temperature and prefers a cool room (Harding and colleagues, 2019). Night noise disturbs sleep and, sustained, raises the risk of cardiovascular disease, which is why the World Health Organization publishes guideline levels for it (WHO, 2009; Basner and colleagues, 2014). Damp and mould in dwellings go with respiratory illness, and the same organisation says so (WHO, 2009). Office workers with windows slept longer at night than workers without them (Boubekri and colleagues, 2014). People who described their homes as cluttered showed flatter daily cortisol rhythms (Saxbe and Repetti, 2010), and clutter went with a weaker sense of home in a large survey (Roster and colleagues, 2016). People judge curved interiors more beautiful than angular ones and are more inclined to enter them (Vartanian and colleagues, 2013), and rooms with higher ceilings prime freer thinking (Meyers-Levy and Zhu, 2007; Vartanian and colleagues, 2015).

The middle of the ladder holds the claims that are supported by observation rather than experiment. Appleton’s prospect–refuge hypothesis, that people prefer places from which they can see without being seen, was proposed from landscape and art rather than measured, and Sommer’s seating studies described where people sit when they can choose rather than what happens if they cannot. The Kaplans’ work on the preference for natural scenes sits with them. These are the sources for the command position’s plausibility. They make it reasonable; they do not make it proven.

Put on one plan, the studies measure seven things, and each has a place in the room where the instrument stood:

Plan 02 The instrumented bedroom: what the studies measure, and where they measure it Seven numbers on one room, each at the place a study put its instrument: the light reaching the pillow, the noise at the street window, the air temperature, the damp at the cold wall, what the other window looks out on, the objects on the floor, and the shape and height of the room. The key names the studies. None of the seven is a compass bearing, and none of the studies asked for one; that is the boundary of the evidence, drawn.

The studies of feng shui itself

Two papers have looked at the tradition directly, and neither tests it. Bonaiuto, Bilotta and Stolfa (2010) set the rules of feng shui beside the findings of environmental psychology and found a partial overlap: the tradition’s preferences for protected positions, for privacy, for orientation to light and for legible circulation are also the discipline’s, and its symbolic rules have no counterpart. Mak and Ng (2005) asked architects in Hong Kong and Sydney to rate a set of feng shui principles for a house, and found that the ones about site, orientation and layout were judged sound design by both groups, and the symbolic ones were not.

There is also the anthropology. Bruun’s fieldwork in China records something no survey captures: practitioners producing different readings of the same house, from different dates, different facings and different schools, and clients choosing among them. A method whose practitioners disagree about the input cannot be right in all of their hands, and this reliability problem comes before any question of validity.

The ledger

Fig. 03 The studies the site leans on most, in one ledger, with what each tested, where and on how many, what it found, and the tradition’s rule it bears on. None of them tested feng shui; each tested a thing the tradition asked for, a window onto trees, a sunny room, a quiet bedroom, an opened window, a cleared floor, and found it worth something, in days, doses, test scores or prices. The sizes matter and are given: Ulrich’s forty-six patients are famous and few, Heschong’s twenty-one thousand pupils are many and observational, the price studies count thousands of sales and measure a belief rather than a house. Read down the last column and the tradition’s room rules appear one by one; read down the fourth and the effects are modest, real and specific, which is what the essay says the evidence is. The numbers are the papers’, rounded; the full citations are on the sources page under research. Studies of the compass schools’ claims are absent because there are none of the kind that would go in this table.

The site has cited a hundred and fifty works by now, and the twelve on the ledger above are the ones its room rules lean on most, set out so that a reader can see what each actually tested and on how many. Read down the ledger and three things show. The first is that none of the twelve tested feng shui; each tested a thing the tradition asked for, a window onto trees, a sunny room, a quiet bedroom, an opened window, a cleared floor, a desk by the light, and found it worth something the paper could count, in days on a ward, doses of a painkiller, points on a reading test, parts per million, a fraction of a price. The second is that the effects are modest and specific: three-quarters of a day, a fifth, a few per cent, a straight line on a graph, and never the transformation the trade promises; the recovering at home guide gives the ward numbers, the student’s desk the classroom ones, the air guide the parts per million and the selling guide the prices, and each says what its numbers do not mean. The third is the sizes. Ulrich’s forty-six patients are the most cited forty-six people in the field and they are forty-six; Heschong’s twenty-one thousand pupils are many and were observed rather than assigned; the price studies count thousands of sales and measure a belief in the buyers rather than a property of the house. A reader who wants to know how much to trust a rule of the site’s can find its study on the ledger, look at the fourth column, and trust it that much. What is not on the ledger is the compass schools, and the essay’s section above says why: there are no studies of the kind that would go in the table, and the site does not fill the gap with the trade’s testimonials.

What has not been studied

The four claims at the foot of the ladder are the ones the tradition is best known for, and nothing has tested them. Bed direction and sleep quality: no trial. Kua directions and health, wealth or marriage: no trial. Flying Star charts and anything at all: no trial. Wealth corners and income: no trial. The absence is not entirely an accident of funding. Each of these claims, as the schools state it, is hard to fail: the facing can be disputed, the period can be disputed, the cures are allowed throughout, the outcomes are broad and the window is twenty years. A claim that cannot fail cannot be supported either, because there is nothing a result could have looked like that would have counted against it.

How the labels work here

Six labels, one claim each, and no averaging. A sentence under the tradition label is what the classical texts and long practice say; it is not made truer by a study on the same page. A sentence under the evidence label names its study and says what the study found and what it did not; it is not extended to cover the tradition’s neighbouring claims. Where a room rule and a finding coincide, as they often do, the site says so under both labels, and the reader can see that the agreement is a coincidence of two routes rather than a proof of one by the other. The methodology page sets out the six labels and the sources policy.

Reading the label

Tradition. The texts and long practice. Read it as a report of what the tradition holds.
School. One lineage’s teaching. Read it as that school’s, and expect another school to differ.
Interpretation. A modern reading, ours or someone else’s, named as such.
Design observation. What people who design rooms have seen, independent of feng shui.
Evidence. A named study, with what it found and what it did not.
Misconception. A popular belief the tradition itself does not support.

A page that carries all six is not contradicting itself. It is showing where the claims on it come from, which is the only honest way to write about a tradition that is partly about rooms and partly about fortune.