How to Predict the Major Product of an Organic Reaction
How to predict the major product of an organic reaction: a six-step procedure through chemo-, regio- and stereoselectivity, with textbook traps.
Add one equivalent of methylmagnesium bromide to an ester and the “obvious” product, a methyl ketone, barely shows up. What you isolate is the tertiary alcohol from two additions, plus unreacted ester. That is the core problem in learning how to predict the major product of an organic reaction: the first product you draw is often the one the reagent was never going to give. This guide is a procedure you can run on any reaction. It goes from functional groups and reagent role through chemo-, regio- and stereoselectivity, then kinetic versus thermodynamic control, and ends with sanity checks.
Before you start: what you need on the page
You need the full scheme, not only the structures. Reagents, equivalents, solvent, temperature and the workup all change the answer. If you are unsure where each item belongs, see what goes above and below the arrow in a reaction scheme. A drawing with no conditions is a structure diagram, and you cannot predict a product from it.
Write down three things before you push any electrons:
- Every functional group on the substrate, including the ones you think are spectators.
- The role of each reagent: nucleophile, electrophile, base, acid, oxidant, reductant or catalyst.
- The stoichiometry: catalytic, one equivalent, or excess.
How to predict the major product of an organic reaction, step by step
Step 1: Classify the reagent, then list everything it can touch
Name what the reagent does before you decide where it does it: a base deprotonates, a hydride donor adds to carbonyls, an organocopper adds to the β-carbon of an enone.
Then go back to your list of functional groups and mark every site the reagent can reach. The product follows from that list. A drawn product that ignores a reactive group on the substrate is a guess.
Step 2: Chemoselectivity, or which group reacts
Chemoselectivity is the preference of a reagent for one functional group over another. Rank the reactive sites by how fast each one reacts with this reagent under these conditions.
The Grignard-and-ester case from the opening is the same logic run over time. The first addition expels methoxide and gives a ketone. That ketone is more electrophilic than the starting ester, so it captures the next equivalent of Grignard faster than the remaining ester does. Doubling the equivalents does not rescue the ketone; it completes the tertiary alcohol. If you want the ketone, change the substrate to a Weinreb amide, whose chelated tetrahedral intermediate holds until workup.
Step 3: Regioselectivity, or where on that group
Once you know which group reacts, ask where. Regioselectivity is the preference for one site of bond formation or cleavage over the other possible ones. Two textbook cases show how reagent choice flips the answer.
Zaitsev versus Hofmann elimination. 2-Bromo-2-methylbutane with sodium ethoxide in ethanol gives mainly 2-methylbut-2-ene, the more substituted (Zaitsev) alkene. With potassium tert-butoxide the major product becomes 2-methylbut-1-ene, the less substituted (Hofmann) alkene. The bulky base has better access to the primary CH3 hydrogens than to the internal CH2. Alkene stability has not changed; what changed is which hydrogen the base can reach.
1,2- versus 1,4-addition. Cyclohex-2-enone has two electrophilic carbons, the carbonyl carbon and the β-carbon. Methyllithium is a hard nucleophile and adds directly to the carbonyl, giving 1-methylcyclohex-2-en-1-ol after workup. Lithium dimethylcuprate (Me2CuLi) is soft, adds conjugatively at the β-carbon, and after protonation of the enolate gives 3-methylcyclohexanone. One methyl group, two different products, and the metal decides which.
Step 4: Kinetic versus thermodynamic control
Ask whether the step is reversible under the conditions. If it is not, the product that forms fastest wins (kinetic control). If it is, the most stable product accumulates (thermodynamic control). Low temperature, a strong non-nucleophilic base and no proton source point to kinetic control. Higher temperature, a weaker base and a proton source that allows equilibration point to thermodynamic control.
The same question settles 1,2- versus 1,4-addition with reversible nucleophiles and many Diels–Alder endo/exo ratios. If you cannot say whether the first step reverses, you cannot yet predict the ratio.
Step 5: Stereoselectivity, including the geometry the mechanism demands
Stereoselectivity is the preferential formation of one stereoisomer over others. Start with the default: an achiral reagent acting on an achiral substrate gives a racemate, however neat the wedge you drew. After that, look for two things.
Face selectivity set by the reagent's size. NaBH4 reduces 4-tert-butylcyclohexanone mainly by axial attack, giving the equatorial alcohol (trans). L-Selectride is too bulky for the axial approach, attacks equatorially, and gives mainly the axial alcohol (cis). Same ketone, opposite major diastereomer.
Geometric requirements of the mechanism. E2 needs the leaving group and the β-hydrogen anti-periplanar, which in a cyclohexane means both axial. Menthyl chloride shows how hard this rule can override Zaitsev. With sodium ethoxide it gives only the less substituted 2-menthene, because in the conformer with an axial chlorine the only anti-periplanar β-hydrogen sits on the carbon away from the isopropyl group. Its diastereomer, neomenthyl chloride, has a choice of anti-periplanar hydrogens and gives mainly the more substituted 3-menthene.
Step 6: Sanity checks before you commit material
- Balance the atoms. Every atom in the product must come from a reactant, and the byproduct should be named: MeOMgBr, Ph3PO, HCl. An atom that appears from nowhere means the mechanism you drew is wrong.
- Check the equivalents against the number of reactive sites. Excess reagent over-reacts, and a deficit leaves starting material. Run the numbers on reagent equivalents and ratios, and put them in a reagent table with the stoichiometry calculator before you weigh anything.
- Read the workup as a reaction. Aqueous acid protonates alkoxides and enolates, hydrolyses acetals and imines, and can dehydrate tertiary alcohols. The product in the flask before workup is not always the product in the vial.
- Name the minor products. A good prediction reads “mostly X, some Y, watch for Z,” not one structure.
Variations and edge cases
Catalysts reshape the selectivity order. CuI turns a Grignard reagent that adds mostly 1,2 to an enone into a 1,4 reagent. Protecting groups change the chemoselectivity list by removing a site. Solvent can change the mechanism, for example by shifting a secondary halide between substitution and elimination. When a catalyst, additive or protecting group appears over the arrow, go back to Step 1 and rebuild the list of reactive sites.
When this procedure runs out
The procedure ranks outcomes by reasoning. It does not replace precedent. For a new substrate class, an unusual catalyst system or a selectivity that depends on fine steric balance, find a literature example on a close analog before you trust your prediction.
Software can help as a second reader. ChemStitch's AI Chat will suggest a product for a reaction you draw, and the answer is labelled AI-suggested. Use it as a cross-check against your own Step 1 to Step 6 reasoning, and treat any disagreement as a prompt to look harder at selectivity, not as the answer. For where model predictions hold up and where they break, see our look at whether AI can predict reaction products reliably.
Run the six steps in order on your next scheme. If your predicted product survives all six, you have a defensible prediction. If it fails one, you found the problem before the TLC plate did.