Yes. Osmosis moves water from a region of high water concentration (low solute concentration) to a region of low water concentration (high solute concentration) across a selectively permeable membrane. The direction is always downhill in water potential, not downhill in solute. That distinction matters because textbook shorthand routinely swaps "water" for "solute" and reverses the sentence without noticing. Water's own chemical potential drives the flow, and it falls as solute concentration rises.

The molecular mechanism behind water's downhill flow

Water potential is the free energy per mole of water, expressed in pressure units. Pure water at atmospheric pressure and a given temperature sits at 0 MPa by convention. Adding solute lowers water potential into negative territory because solute particles occupy space, hydrate, and reduce the fraction of water molecules able to escape into vapor or cross a membrane. A 0.1 molal solution of a non-ionizing solute sits near −0.25 MPa at 20 °C, and seawater near −2.4 to −2.5 MPa.

Water crosses membranes through two routes. It diffuses through the lipid bilayer itself, and it streams through aquaporins, channel proteins that pass roughly 3 billion water molecules per second each in some measurements. Both routes are passive. Neither consumes ATP. The driving force is the gradient in water potential, which has a pressure component and a solute (osmotic) component.

Aquaporins raise permeability without changing direction. A membrane with ten times more aquaporin still moves water down the same gradient, just faster. This is why some tissues equilibrate in seconds and others take hours. The gradient sets the arrow. The proteins set the speed.

Two proteins do change the arrow's consequences. Osmotic stress activates kinases that regulate aquaporin trafficking, and cells adjust solute content through transporter proteins to defend volume. These are responses to the gradient, not reversals of it. Active transport of ions can build a gradient that water then follows, which is the setup for every absorptive epithelium in the body.

Quantified context: numbers that anchor the rule

Osmotic pressure follows Π = iMRT, where i is the van't Hoff factor, M is molarity, R is the gas constant, and T is temperature in kelvin. For a non-electrolyte at body temperature, 1 milliosmole per kilogram of water generates roughly 19.3 mmHg of osmotic pressure, which converts to about 0.00257 MPa. Plasma at 290 mOsm/kg therefore holds about 7.4 atm, or 0.75 MPa, of osmotic pressure against pure water.

Setting Solute concentration Osmotic pressure Direction of water flow
Pure water 0 mOsm/kg 0 MPa Reference point
Typical mammalian plasma ~290 mOsm/kg ~0.75 MPa Into plasma from pure water
Red blood cell interior ~300 mOsm/kg ~0.77 MPa Into cell if placed in pure water
Seawater ~1,000 mOsm/kg ~2.4–2.5 MPa Out of a marine fish's body
Freshwater pond ~5–20 mOsm/kg near 0.01–0.05 MPa Into a freshwater fish
Sucrose used in food preservation ~2,000–4,000 mOsm/kg >5 MPa Out of microbial cells

Those figures explain why a marine teleost drinks seawater continuously and excretes concentrated salt, while a freshwater fish never drinks and produces copious dilute urine. The gradient direction is fixed. The organism's plumbing is what differs.

Preservation thresholds are practical numbers. Solutions above roughly 20 percent sucrose or 15 percent salt hold water activity below about 0.85, low enough that most bacteria cannot pull water in. The osmotic gradient removes water from microbial cytoplasm faster than the cell can defend. This is osmosis in a jar, and it follows the same equation as a kidney.

Edge cases where the simple sentence fails

The first failure is semantics. Osmosis is defined as water movement, so "high to low concentration" must refer to water. In solute terms, water moves from low solute concentration to high solute concentration. Both sentences describe the same event. Mixing them produces the contradiction that confuses students.

The second failure is turgor pressure. In a plant cell, water enters until internal hydrostatic pressure offsets the osmotic gradient. At equilibrium, net flow stops even though solute concentrations remain unequal. Pressure potential has risen to cancel the solute potential. The gradient did not vanish; the total water potential did. Cells with rigid walls reach this state at turgor pressures of roughly 0.5 to 1.0 MPa in many herbaceous tissues.

The third failure is active transport. The loop of Henle reverses local osmotic expectations by pumping ions with ATP, creating a hypertonic medulla. Water then follows gradients in the descending limb and is withheld in the ascending limb. The water movement is still passive. The gradient that drives it was built by energy expenditure elsewhere.

The fourth failure involves unstirred layers. Adjacent to any membrane sits a layer of fluid that diffusion must cross, and its thickness can limit measured water flux. A membrane with enormous aquaporin density can still appear modestly permeable because the unstirred layer, not the pore, becomes rate-limiting. This is the professional caveat most explainers omit: measured osmotic permeability can understate true membrane permeability, and comparisons across studies must control for stirring and layer thickness.

Misconceptions that survive into professional writing

Misconception one: osmosis requires a membrane, but the membrane does not have to be biological. Dialysis tubing, reverse-osmosis filters, and cellophane all produce it. The requirement is selective permeability, meaning solute passage is restricted relative to water.

Misconception two: osmosis is the same as diffusion. Osmosis is a special case of diffusion constrained by a semipermeable barrier. Diffusion of a solute down its own gradient proceeds without any barrier. Water's movement in osmosis is bulk flow driven by chemical potential, and in channels it can be collective rather than independent molecular hops.

Misconception three: cells shrink in salt because salt "pulls" water. Salt does not attract water through a membrane. Solute lowers water's chemical potential on its side, and water moves to equalize that potential. The phrase "salt sucks" is a mnemonic, not a mechanism.

Counterintuitive detail: a cell can swell in a solution that is hypertonic to pure water but hypotonic to its own interior. Direction is relative, and the reference compartment is the cell, not the room. Physiologists compare every external solution to intracellular osmolality, near 300 mOsm/kg in mammalian cells.

Warning: assuming "high to low" means high solute to low solute reverses the water direction; water goes toward the solute. Warning: applying osmotic predictions to cells with active transporters without accounting for ion pumping will mispredict volume. Warning: comparing osmotic permeability values across studies without noting unstirred-layer corrections can mislead by large margins. Warning: treating aquaporin abundance as sufficient to predict flow ignores the gradient magnitude and membrane area. Warning: using seawater figures as a universal reference misjudges freshwater and estuarine organisms.

Practical actions: using the rule correctly

Start by identifying the compartments. Write down the solute concentration or osmolality on both sides of the membrane in the same units. Then set the direction from high water potential to low, which means from low osmolality to high.

Estimate the magnitude with Π = iMRT before reasoning qualitatively. A 0.9 percent sodium chloride solution is isotonic to plasma at about 308 mOsm/L because sodium chloride dissociates into two particles. A 5 percent dextrose solution is also near isotonic at roughly 278 mOsm/L, but it behaves hypotonic after glucose is metabolized, which is why it is described as isotonic in the bag and hypotonic in the body.

Check for an offsetting pressure. Plant cells and bacteria generate turgor, and animal cells regulate volume. If a predicted flow does not occur, look for a pump, a channel, or a hydrostatic pressure that cancels the osmotic term.

When designing a protocol, hold temperature and stirring constant if you compare permeability. Report osmolality rather than percent concentration when solutes differ in dissociation. Report the membrane area and the unstirred-layer conditions.

For clinical or laboratory decisions involving intravenous fluids, dialysis, or cell culture media, confirm composition against the manufacturer's specification and institutional guidance. The direction of osmosis is reliable. The exact volume change is not, because it depends on permeability coefficients and regulation that vary by tissue and time.

Limits of what this rule can predict

The rule predicts direction reliably and magnitude crudely. It cannot predict how much water moves without permeability data, membrane area, and time. It cannot predict whether a cell survives the volume change. It cannot predict the behavior of a tissue with heterogeneous aquaporin distribution, and it cannot predict the steady state when active transport and metabolism are running.

Reverse osmosis is the engineering proof that direction is not a law of nature but a consequence of water potential. Apply external pressure above the osmotic pressure, and water flows from high solute to low solute against the natural direction. A seawater desalination unit must exceed roughly 2.4 to 2.5 MPa, and modern units operate near 5.5 to 7 MPa to achieve practical recovery. The gradient can be overridden by pressure, which is precisely what the water-potential framework predicts.

The durable answer, then, is that osmosis carries water from high water concentration to low water concentration. It moves toward higher solute concentration and toward lower water potential. Any statement that puts solute on the wrong side of the sentence is a labeling error, not a new finding. The mechanism is passive and universal, and its exceptions are optical, not physical.

For mechanisms at the level of membrane biophysics, the standard reference remains the textbook treatment of water potential and aquaporin physiology, and the osmotic pressure relation traced to van't Hoff is reproduced in every physiology text. The American Physiology Society and the National Institutes of Health physiology resources carry consistent treatments of these definitions.