Finding the hybridization of a central atom comes down to one core step: count the number of electron groups around it. Each bond (single, double, or triple counts as one group) and each lone pair counts as one group. Two groups mean sp hybridization, three mean sp², four mean sp³, and so on. That single counting exercise handles the vast majority of molecules you will encounter, but the interesting chemistry starts where the simple count meets lone pairs, expanded octets, and atoms that refuse to follow the textbook angles.
The Steric Number Method
The steric number is the total count of electron groups around a central atom. An “electron group” is any region of electron density pointed away from the center: a single bond, a double bond, a triple bond, or a lone pair. A double bond between carbon and oxygen still counts as just one group because both pairs of shared electrons sit in roughly the same spatial direction. A triple bond counts as one group for the same reason. Only lone pairs add groups without adding visible bonds in a structural formula.
Once you have the steric number, the hybridization follows directly:
- Steric number 2: sp hybridization, linear geometry (180° between groups)
- Steric number 3: sp² hybridization, trigonal planar arrangement (120°)
- Steric number 4: sp³ hybridization, tetrahedral arrangement (109.5°)
- Steric number 5: sp³d hybridization, trigonal bipyramidal arrangement
- Steric number 6: sp³d² hybridization, octahedral arrangement
So for water, the oxygen has two bonds to hydrogen and two lone pairs: steric number 4, sp³. For carbon dioxide, the carbon has two double bonds and no lone pairs: steric number 2, sp. For ammonia, the nitrogen has three bonds and one lone pair: steric number 4, sp³. The method works the same way every time. You are not counting atoms or bonds individually. You are counting directions in which electron density points away from the central atom.
Drawing the Lewis Structure First
You cannot count electron groups without a correct Lewis structure, and this is where most mistakes happen. If you draw the wrong number of lone pairs on the central atom, you will get the wrong steric number and therefore the wrong hybridization. A common error is forgetting lone pairs on atoms like nitrogen, oxygen, and sulfur, or putting too many on atoms that actually participate in double bonds.
For molecules with resonance structures, the hybridization is usually the same across all valid Lewis structures. Take the nitrate ion: no matter which oxygen you draw the double bond to, nitrogen always has three groups around it (one double bond and two single bonds, or three equivalent bonds in the resonance hybrid). The steric number stays 3 and the hybridization stays sp² regardless of which resonance form you sketch. If two resonance structures give you different steric numbers, that is a strong sign one of your Lewis structures has an error.
For charged species, remember that formal charge changes the electron count. A positively charged nitrogen, for instance, may have lost a lone pair compared to its neutral form, dropping its steric number. The ammonium ion (NH₄⁺) has nitrogen bonded to four hydrogens with no lone pairs: steric number 4, sp³. Neutral ammonia also has steric number 4, but one of those groups is a lone pair rather than a bond.
Why Lone Pairs Matter So Much
Lone pairs occupy hybrid orbitals just like bonding pairs do. The critical difference is that lone pairs are held by only one nucleus instead of two, so they spread out more and compress the angles between bonding pairs. This is why water is not a straight line even though it has only two bonds. The two lone pairs on oxygen push the two O–H bonds closer together, squeezing the H–O–H angle from the ideal tetrahedral 109.5° down to about 104.5°.
This compression does not change the hybridization assignment. Water is still sp³ because oxygen still has four electron groups. But it does change the molecular shape you would report. The hybridization tells you the arrangement of all electron groups; the molecular geometry describes only where the atoms sit. An sp³ atom with no lone pairs gives a tetrahedral shape. With one lone pair, you get a trigonal pyramidal shape. With two lone pairs, a bent shape. All three are sp³, but they look completely different as molecules.
Beginners sometimes confuse molecular geometry with hybridization. A bent molecule is not always sp² just because it looks like a fragment of a trigonal plane. Check the lone pairs. If the central atom has two bonds and two lone pairs (like water), it is sp³ and bent. If it has two bonds and one lone pair (like sulfur dioxide), it is sp² and bent. Same shape label, different hybridization, different bond angles.
Molecules With More Than Four Groups
Atoms in the third row of the periodic table and below can accommodate five or six electron groups. Phosphorus pentachloride (PCl₅) has five bonds and no lone pairs around phosphorus: steric number 5, which is assigned sp³d hybridization in the traditional model. Sulfur hexafluoride (SF₆) has six bonds: steric number 6, sp³d² hybridization. These expanded-octet cases follow the same counting rule, just with higher numbers.
The underlying explanation for why these molecules exist, however, has shifted over the decades. The traditional textbook story invokes d-orbitals on the central atom mixing in to provide the extra room. Modern computational work tells a different story. Research into hypervalent bonding has shown that d-orbital participation in these molecules is minimal. Instead, the bonding in species like PCl₅ and SF₆ is better described as highly polar bonds with significant ionic character, or through multi-center bonding models. The electronic structure of these molecules has been interpreted through the three-center, four-electron bonding model originally proposed by Rundle and Pimentel, as well as through charge-transfer and hypercoordinate frameworks.1PubMed Central. Symmetry of three-center, four-electron bonds
For practical purposes, the steric-number-to-hybridization lookup still gives you the right geometry and the right prediction for molecular shape. The labels sp³d and sp³d² remain useful as bookkeeping tools even if the physics underneath is more nuanced than the simple “d-orbitals mix in” story. If you are taking a general chemistry or organic chemistry course, the counting method works. If you are doing computational chemistry, you would use molecular orbital theory rather than hybridization labels for these species.
When Bond Angles Deviate From the Ideal
The steric number method predicts ideal angles: 109.5° for sp³, 120° for sp², 180° for sp. Real molecules rarely hit these numbers exactly. One reason is lone-pair repulsion, as discussed with water. Another, more subtle reason involves the electronegativity of the atoms bonded to the center.
A principle known as Bent’s rule captures this effect: atoms direct hybrid orbitals with more p-character toward more electronegative substituents and orbitals with more s-character toward less electronegative ones. Research has confirmed that this classic correlation between electronegativity and hybridization holds broadly across main-group elements, not just carbon compounds, though it also has limits.2PubMed. Hybridization trends for main group elements and expanding the Bent’s rule beyond carbon: more than electronegativity More recent work has extended the principle to explain conformational stability differences in complex molecules like acetylcholine, suggesting the rule has wide-ranging utility beyond simple textbook examples.3Malaysian Journal of Fundamental and Applied Sciences. Electronegativity Effects on Conformational Stability Using Bent’s Rule: From Simple Molecules to Acetylcholine
What this means in practice is that the hybridization label (sp³, sp², etc.) gives you the approximate geometry, but the actual bond angles can shift by several degrees depending on what is attached to the central atom. Fluorine substituents, being extremely electronegative, pull more p-character into the bonds they form, leaving more s-character in the remaining bonds. This causes measurable angle distortions. You would not change the hybridization label because of these shifts, but you should not be surprised when a measured bond angle is a few degrees off from the textbook ideal.
Reactive Intermediates and Unusual Species
Carbocations, carbanions, and free radicals are short-lived species that show up constantly in organic reaction mechanisms, and their hybridization is not always obvious from the usual rules.
A carbocation (a carbon with a positive charge and only three bonds) has steric number 3 and is sp² hybridized. The empty p-orbital that remains unhybridized sits perpendicular to the plane of the three bonds. This geometry is directly responsible for how carbocations react: nucleophiles can attack from either side of that empty orbital.
A carbanion (a carbon carrying a negative charge and a lone pair) formally has four electron groups: three bonds plus one lone pair. By the counting method, that gives sp³. In practice, some carbanions are better described as sp² with the lone pair in a p-orbital, especially when the negative charge is stabilized by resonance with an adjacent double bond or aromatic ring. The hybridization depends on the specific molecule, not just the charge.
Free radicals have an unpaired electron occupying an orbital. A methyl radical (·CH₃) is nearly planar, suggesting sp² hybridization with the odd electron in a p-orbital, though computational studies show the geometry is not perfectly flat. Larger radicals can be pyramidal and closer to sp³. The point is that reactive intermediates are where the simple counting rules start to need context. You can still use the steric number as a starting point, but the answer sometimes depends on what the molecule is doing electronically rather than just how many groups you can count.
Transition Metal Complexes
When you move beyond main-group chemistry into transition metals, hybridization gets used differently. A coordination compound like [Co(NH₃)₆]³⁺ has six ligands arranged octahedrally around cobalt. In the hybridization-based model, this is described as d²sp³ hybridization (or sp³d² depending on the textbook convention). Hybridization-based theory has been used to explain the geometry, bonding, and reactivity of many transition-metal complexes, sometimes drawing parallels to simpler main-group systems through concepts like the isolobal analogy.4Inorganic Chemistry. Understanding Transition-Metal Complexes Using Hybridization-Based Theory
That said, most modern inorganic chemistry courses rely on crystal field theory or ligand field theory rather than hybridization to explain transition-metal bonding. These approaches account for color, magnetism, and spectroscopic properties in ways that hybridization alone cannot. If you are asked to find the hybridization of a transition-metal center, the steric number approach still works for predicting geometry (six groups = octahedral, four groups = tetrahedral or square planar), but it will not tell you much about the electronic properties that actually matter in d-block chemistry.
Square planar complexes are a particularly tricky case. Four groups around a center normally predict tetrahedral geometry (sp³), but many d⁸ metal ions like Pt²⁺ and Pd²⁺ adopt a square planar shape instead. The hybridization is labeled dsp² in the traditional scheme. You cannot predict this from steric number alone; you need to know the metal’s electron count and the strength of the ligand field. This is one area where the simple counting method genuinely fails and more advanced theory is required.
Where the Simple Model Hits Its Limits
Hybridization is a model, not a physical measurement. No instrument reads out “sp²” when you point it at a molecule. The concept was introduced by Linus Pauling in the 1930s as a way to reconcile the known shapes of molecules with the atomic orbitals of isolated atoms. Modern computational methods have repeatedly confirmed that Pauling’s qualitative picture of directional hybrid orbitals holds up well across a wide range of quantum chemistry approaches.5PubMed Central. Pauling’s Conceptions of Hybridization and Resonance in Modern Quantum Chemistry The model is robust and useful. But it does have boundaries.
One boundary that chemists have debated is whether it even makes sense to assign a hybridization to certain atoms. Some researchers have argued that hybridization labels should not be applied to lone pairs and bonds within a molecule as though they are fixed properties, since the orbitals are perturbed during bond formation. Under this critique, discussing hybridization for a terminal atom in a polyatomic molecule, or for either atom in a diatomic molecule, is not strictly justifiable.6Quimica Nova. HYBRID ATOMIC ORBITALS IN ORGANIC CHEMISTRY. PART 2: CRITIQUE OF PRACTICAL ASPECTS This does not mean the model is useless, but it does mean the labels are approximations rather than exact descriptions of where electrons actually are.
Another limitation is that hybridization says nothing about bond strength or reactivity on its own. Knowing that a carbon is sp² tells you the geometry is trigonal planar and the bond angles are roughly 120°, but it does not tell you how reactive that carbon is, how stable the molecule is, or what will happen when it encounters a reagent. For those questions you need to think about the entire molecular orbital picture, not just the hybridization of one atom.
A Quick-Reference Walkthrough
If you want a reliable procedure you can apply to any molecule on an exam or homework problem, here it is step by step:
- Draw the Lewis structure: Get the total valence electrons right, distribute bonds and lone pairs, and check formal charges.
- Identify the central atom: This is usually the least electronegative atom (hydrogen is never the center; it can only form one bond).
- Count electron groups: Every bond (single, double, or triple) counts as one. Every lone pair on the central atom counts as one.
- Match to hybridization: 2 groups = sp, 3 = sp², 4 = sp³, 5 = sp³d, 6 = sp³d².
- Determine molecular shape: Subtract the lone pairs from the total to get the arrangement of atoms only.
This procedure handles methane, carbon dioxide, water, ammonia, boron trifluoride, sulfur hexafluoride, phosphorus pentachloride, and the vast majority of molecules that appear in general and organic chemistry courses. Where it gets tricky is with species that have resonance structures where the steric number could appear ambiguous, with transition metals, and with reactive intermediates whose geometry depends on electronic stabilization rather than simple counting. In those cases, the counting method still gives you a reasonable starting guess, but you may need additional context to pin down the final answer.
Common Mistakes and How to Avoid Them
The single most frequent error is counting a double bond as two groups instead of one. A double bond has two pairs of shared electrons, but they both point in the same direction relative to the central atom. Carbon in formaldehyde (H₂C=O) has three groups (two C–H bonds and one C=O double bond), not four. That makes it sp², not sp³.
The second most common mistake is forgetting lone pairs entirely. Nitrogen in ammonia has four groups, not three. If you only count the three N–H bonds, you would incorrectly assign sp² and predict a flat molecule. Ammonia is pyramidal precisely because that lone pair is there, occupying one corner of the tetrahedron.
A subtler error is trying to assign hybridization to hydrogen or to terminal atoms in general. Hydrogen has only one orbital that matters (the 1s orbital), so hybridization does not apply. Terminal fluorine atoms bonded to a central atom do not need a hybridization label to understand the molecule’s geometry. The hybridization question is about the central atom, and that is where your attention should stay.
Finally, do not confuse electron-group geometry with molecular geometry. Both sp³ carbon in methane and sp³ nitrogen in ammonia have electron groups arranged tetrahedrally. But methane’s molecular shape is tetrahedral (all four groups are bonds), while ammonia’s molecular shape is trigonal pyramidal (one group is a lone pair). The hybridization is the same; the visible shape is not. Keeping that distinction clear prevents a whole category of errors on exams and in lab reports.