Finding a molecule’s geometry starts with a surprisingly simple idea: count the groups of electrons around the central atom, figure out how they push each other apart, and the shape follows. This approach, rooted in a model called VSEPR (valence-shell electron-pair repulsion), handles most everyday molecules reliably. But for more complex or exotic species, and for confirming what theory predicts, chemists turn to experimental techniques like X-ray diffraction and spectroscopy, or to computational methods that simulate electron behavior directly.
Count the Electron Groups Around the Central Atom
The most widely taught method for predicting molecular geometry is VSEPR, and it works because electrons repel each other. Every group of electrons around a central atom, whether it forms a bond to another atom or sits as a non-bonding lone pair, wants to get as far away from the other groups as possible. The resulting arrangement of electron groups is called the electron-pair geometry, and the molecular geometry is what you see when you look at only the positions of the atoms, ignoring the invisible lone pairs.
To use this method in practice, you follow a short sequence of steps. First, draw the Lewis structure of the molecule. Identify the central atom (usually the least electronegative atom that isn’t hydrogen). Then count the number of electron groups attached to it. A single bond, a double bond, and a triple bond each count as one group, because all the electrons in a given bond point in the same direction and repel other groups as a unit. Lone pairs also count as one group each.
Two electron groups arrange themselves in a straight line (linear). Three spread into a flat triangle (trigonal planar). Four point toward the corners of a tetrahedron. Five form a trigonal bipyramid. Six create an octahedron. These are the electron-pair geometries, and they represent the ideal arrangements that minimize repulsion. The molecular geometry, the shape you actually report, depends on how many of those positions are occupied by atoms versus lone pairs.
How Lone Pairs Change the Picture
Lone pairs take up more angular space than bonding pairs because they are held closer to the central atom and spread out more. This extra bulk squeezes the bonding pairs closer together, distorting the ideal angles. Water is the classic example: oxygen has four electron groups (two bonds to hydrogen, two lone pairs), so the electron-pair geometry is tetrahedral, but the molecular geometry you report is “bent” because you only see the two hydrogen atoms. The bond angle shrinks from the ideal tetrahedral angle of about 109.5 degrees down to roughly 104.5 degrees.
Ammonia works the same way. Nitrogen has four electron groups (three bonds, one lone pair), giving a tetrahedral electron-pair geometry but a pyramidal molecular shape. The bond angle compresses to about 107 degrees. Every time you swap a bonding pair for a lone pair, the shape name changes even though the underlying electron arrangement stays the same.
In molecules with five or six electron groups, lone pairs preferentially occupy certain positions. In a trigonal bipyramid, lone pairs sit in the equatorial plane where they have more room. This leads to shapes like seesaw (one lone pair), T-shaped (two lone pairs), and linear (three lone pairs). Computational studies of molecules with central atoms from the heavier main-group elements confirm that lone-pair-driven distortions generally follow these patterns, though the degree of distortion varies depending on the size and electronegativity of the surrounding atoms.1PubMed. DFT-based pseudo-Jahn-Teller coupling studies on the steric and energetic lone pair effect of four- and five-coordinate halide molecules and complexes with central ions from the fifth, sixth, and seventh main groups
A Quick Reference for Common Shapes
Once you have counted your electron groups and identified how many are lone pairs, you can look up the resulting shape. Here are the most frequently encountered molecular geometries, organized by the number of bonding groups and lone pairs on the central atom:
- Linear: two bonding groups, no lone pairs. Examples include carbon dioxide and beryllium chloride. Bond angle is 180 degrees.
- Trigonal planar: three bonding groups, no lone pairs. Boron trifluoride is a textbook case. Bond angles are 120 degrees.
- Bent (from three groups): two bonding groups plus one lone pair. Sulfur dioxide fits here, with a bond angle slightly less than 120 degrees.
- Tetrahedral: four bonding groups, no lone pairs. Methane is the prototype. Bond angles are 109.5 degrees.
- Trigonal pyramidal: three bonding groups plus one lone pair. Ammonia, with bond angles around 107 degrees.
- Bent (from four groups): two bonding groups plus two lone pairs. Water, with a bond angle near 104.5 degrees.
- Trigonal bipyramidal: five bonding groups, no lone pairs. Phosphorus pentachloride is the standard example.
- Octahedral: six bonding groups, no lone pairs. Sulfur hexafluoride.
Seesaw, T-shaped, square pyramidal, and square planar geometries all arise from these same electron-pair frameworks when one or more positions hold lone pairs instead of bonds. The system is internally consistent: once you know the total number of electron groups, you know the electron-pair geometry, and once you know how many are lone pairs, you know the molecular shape.
Hybridization as a Complementary Lens
VSEPR tells you the shape, but it doesn’t explain why bonds form in certain directions. That explanation comes from hybridization, the idea that atomic orbitals on the central atom blend together to form new hybrid orbitals oriented toward the bonding partners. A carbon atom with four groups uses four equivalent hybrid orbitals pointing toward the corners of a tetrahedron. A carbon with three groups uses three hybrids arranged in a trigonal plane, with the leftover unhybridized orbital available for a second bond in a double bond.
Hybridization and VSEPR almost always agree on the geometry, because they are describing the same physical reality from different angles. But hybridization also explains structural and reactivity trends that VSEPR cannot. The degree of blending between different types of orbitals affects bond lengths, bond strengths, and the acidity or basicity of atoms in a molecule. These effects are pervasive in organic and main-group chemistry and help explain why molecules with the same VSEPR shape can behave very differently in reactions.2Journal of Physical Organic Chemistry. Orbital hybridization: a key electronic factor in control of structure and reactivity
When the Simple Model Falls Short
VSEPR handles most small molecules well, but it has blind spots. Transition-metal complexes often adopt geometries that VSEPR doesn’t predict, because the metal’s partially filled inner orbitals interact with the surrounding ligands in ways that the simple repulsion model doesn’t account for. Crystal-field and ligand-field theories handle these cases, and the resulting geometries (square planar, for instance, in many four-coordinate complexes of platinum and palladium) can seem counterintuitive from a VSEPR perspective.
Some lone pairs are also less geometrically active than expected. In heavy elements, the lone pair can become “stereochemically inert,” meaning it doesn’t push bonds out of position the way lighter-element lone pairs do. Studies of gold anion compounds, for example, found that the lone pair on the gold anion barely perturbs the structure at all, even under electrical or mechanical stress.3PubMed. On the stereochemical inertness of the auride lone pair: ab initio studies of AAu (A = K, Rb, Cs) This effect shows up across several heavy main-group elements and means that the simple rule “lone pairs compress bond angles” doesn’t always hold as you move down the periodic table.
Another complication is that some molecules don’t hold a single fixed geometry. Pentavalent phosphorus compounds, for example, are fluxional, meaning the atoms constantly swap positions between the axial and equatorial sites of the trigonal bipyramid. Computational work has confirmed that these rearrangements happen through a well-defined pathway called Berry pseudorotation, where the molecule passes through a square pyramidal intermediate on its way from one arrangement to another.4PubMed Central. Stereomutation of pentavalent compounds: validating the Berry pseudorotation, redressing Ugi’s turnstile rotation, and revealing the two- and three-arm turnstiles If you measure the geometry of such a molecule, what you capture depends on the timescale of your measurement. A fast technique might catch the molecule in one arrangement; a slower one might see an average of several.
Measuring Geometry Experimentally
Predicting a shape on paper is one thing; confirming it in the lab is another. Several experimental techniques can pin down bond lengths and angles with high precision, each with its own strengths.
X-ray crystallography is the workhorse for solid-state structures. When X-rays pass through a crystal, they scatter off the electrons surrounding each atom. The pattern of scattered rays can be mathematically reconstructed into a three-dimensional map of where the electrons are densest, revealing the positions of the atoms. Heavier atoms scatter X-rays more strongly and are therefore easier to locate precisely.5Advanced Drug Delivery Reviews. Engineering of pharmaceutical cocrystals, salts and polymorphs: Advances and Challenges – Section: 2.1.2 Single crystal X-ray diffraction This technique has determined the structures of hundreds of thousands of molecules, from simple salts to sprawling organic frameworks.
For gas-phase molecules, rotational spectroscopy offers a complementary approach. Molecules tumble in the gas phase at rates that depend on their mass distribution. By measuring the frequencies at which a molecule absorbs microwave radiation as it rotates, researchers can extract moments of inertia and, from those, precise atom positions. Modern broadband instruments analyze all possible single-isotope substitutions of a molecule, building up the structure atom by atom with bond-length accuracy down to about 0.01 angstroms.6Chemical Physics Letters. Broadband Fourier transform rotational spectroscopy for structure determination: The water heptamer
Nuclear magnetic resonance (NMR) spectroscopy is especially powerful for molecules in solution, including large biological molecules that resist crystallization. The nuclear Overhauser effect, which measures how nearby nuclei influence each other’s magnetic relaxation, provides distance measurements between pairs of atoms. Exact versions of this technique can resolve inter-atomic distances to within about 0.1 angstroms, producing protein and RNA structures of comparable quality to those derived from more traditional methods.7PubMed Central. Advances in the exact nuclear Overhauser effect – Section: 2. eNOE protocol and advantages over conventional NOE
For very large biological assemblies, cryo-electron microscopy has become a transformative tool. Samples are flash-frozen in a thin layer of ice, and images taken from many angles are computationally combined into a three-dimensional reconstruction. This method can now achieve near-atomic resolution and captures molecules in conditions much closer to their natural environment than crystallography does.8PubMed Central. Unravelling biological macromolecules with cryo-electron microscopy It is particularly valuable for dynamic complexes that adopt multiple shapes, since the images can be sorted into groups representing different conformations.9PubMed. Cryo-electron microscopy for structural analysis of dynamic biological macromolecules
Vibrational spectroscopy rounds out the toolkit. Infrared and Raman spectroscopy probe how a molecule vibrates, and the pattern of active vibrations is dictated by the molecule’s symmetry. By analyzing which vibrational modes appear in the infrared spectrum versus the Raman spectrum, researchers can narrow down or confirm the symmetry group of a crystal or molecule.10Journal of Raman Spectroscopy. The use of Raman and infrared spectroscopy in determining the space symmetry group among the groups with the same rules of systematic absence in the diffraction patterns: Some basic principles and applications
Computational Prediction
When experiments are impractical or when you want to explore molecules that haven’t been made yet, computational chemistry fills the gap. The most common approach for geometry prediction is density functional theory (DFT), which approximates how electrons distribute themselves around nuclei and then searches for the arrangement of atoms that gives the lowest total energy. That lowest-energy arrangement is the predicted geometry.
The optimization process typically works in internal coordinates, meaning it adjusts bond lengths, bond angles, and dihedral angles iteratively until the forces on all atoms are essentially zero. Modern algorithms handle molecules ranging from small organic compounds to large coordination complexes, though very floppy molecules with many rotatable bonds require specialized coordinate systems to converge efficiently.11PubMed Central. Geometry optimization in density functional methods
Computational methods also excel at finding transition states and short-lived intermediates that experiments can’t easily isolate. If you want to know the geometry of a reaction intermediate that exists for only a few femtoseconds, computation is often the only realistic option. The results are only as good as the level of theory used, but for most main-group molecules, modern DFT methods produce bond lengths and angles that match experiment to within a few hundredths of an angstrom.
The Same Molecule Can Have Different Shapes in Different Settings
One underappreciated subtlety is that a molecule’s geometry is not a single fixed answer. It depends on the environment. A molecule floating free in the gas phase adopts whatever shape minimizes its own internal energy. Pack that same molecule into a crystal, and the forces exerted by neighboring molecules can push bonds into slightly different angles or lengths. The simplest phosphonium ylide provides a striking example: its gas-phase structure, determined by electron diffraction and confirmed by calculations, differs from the structure found by X-ray crystallography in the solid state.12Angewandte Chemie International Edition. Differences Between Gas-Phase and Solid-State Molecular Structures of the Simplest Phosphonium Ylide, Me3P=CH2
Solvent effects can also matter. Polar solvents stabilize charge-separated forms of a molecule, and that redistribution of electron density can shift bond angles and lengths measurably. When someone asks “what is the geometry of this molecule,” the honest answer sometimes includes “it depends on where the molecule is.” For most common molecules the differences are small, but for polar or highly flexible species they can be significant enough to change how the molecule interacts with its surroundings.
Why Getting the Shape Right Matters
Molecular geometry isn’t just an academic exercise. In drug design, the shape of a small molecule determines whether it fits into the binding pocket of a target protein. Shape complementarity between a drug candidate and its target is often the single most important factor governing how strongly the drug binds, and by extension how well it works. Computational tools that evaluate shape matching are now standard in pharmaceutical research, used to screen millions of candidate compounds before any are synthesized.13PubMed Central. The importance of discerning shape in molecular pharmacology
Chirality is another domain where geometry is decisive. Two mirror-image forms of the same molecule have identical bond lengths and angles but different three-dimensional arrangements, and they can have dramatically different biological effects. One form of a chiral drug might be therapeutic while its mirror image is inactive or even harmful. The understanding that molecular shape matters this much traces back to van’t Hoff’s proposal of the tetrahedral carbon in the 1870s, an insight that placed symmetry and dissymmetry at the center of chemical thinking and launched the entire field of stereochemistry.14PubMed Central. Optical signatures of molecular dissymmetry: combining theory with experiments to address stereochemical puzzles
Materials science depends on geometry as well. The angle at which silicon and oxygen atoms connect in a silicate determines whether you get quartz, glass, or a zeolite with molecule-sized pores useful for catalysis. Polymer properties like flexibility and melting point trace back to the geometry around each carbon-carbon bond in the chain. Even in atmospheric chemistry, the bent shape of water and the linear shape of carbon dioxide explain key differences in how those gases absorb infrared radiation and contribute to the greenhouse effect. Shape, in short, is where the abstract world of electrons and orbitals meets the concrete world of properties you can see, feel, and measure.