A 9-Number Lock That Has Exactly One Key
A 3x3 matrix can rotate, stretch, and shear three-dimensional space all at once. Undoing that transformation — sending every point back exactly where it came from — is what the inverse does, and a $3 \times 3$ matrix either has exactly one such undo or none at all.
That reversibility is the whole point. Solve a system of three equations in three unknowns, undo a 3D rotation in graphics, decode a Hill cipher — each one is the same move: multiply by the inverse and the transformation unwinds. The hard part is computing it, and a 3x3 has just enough moving parts to punish a careless step.
What Is the Inverse of a 3x3 Matrix?
The inverse of a 3x3 matrix $A$ is the unique $3 \times 3$ matrix $A^{-1}$ that satisfies
$$A A^{-1} = A^{-1} A = I,$$
where $I$ is the $3 \times 3$ identity matrix (ones down the diagonal, zeros elsewhere). Multiplying a matrix by its inverse returns the identity — the matrix equivalent of multiplying a number by its reciprocal to get 1.
A 3x3 matrix has an inverse if and only if its determinant is non-zero. A matrix with a non-zero determinant is called invertible (or non-singular); one whose determinant is zero is a singular matrix and has no inverse at all. This is the same invertibility condition that governs every square matrix — the general inverse-of-a-matrix concept covers the 2x2 shortcut and the broader theory; here the focus is the 3x3 case, where the arithmetic gets real.
What Is the Determinant of a 3x3 Matrix?
Before any inverse, compute the determinant — if it comes out zero, stop, because no inverse exists. For
$$A = \begin{bmatrix} a & b & c \ d & e & f \ g & h & i \end{bmatrix},$$
the determinant expands along the first row:
$$\det A = a(ei - fh) - b(di - fg) + c(dh - eg).$$
Each bracket is a $2 \times 2$ determinant — the minor of the leading entry — and the signs alternate $+,,-,,+$. That alternation is not decoration; it is the cofactor sign pattern showing up early, and it returns in full force when you build the adjugate.
What Is the Inverse of a 3x3 Matrix Formula?
The adjugate formula is the standard route:
$$A^{-1} = \frac{1}{\det A},\text{adj}(A),$$
where $\text{adj}(A)$ is the adjugate — the transpose of the cofactor matrix. Reading it as a procedure gives four steps, in order:
Compute $\det A$. If it is zero, there is no inverse — stop here.
Build the cofactor matrix. For each entry, delete its row and column, take the determinant of the remaining $2 \times 2$ (its minor), and attach the sign $(-1)^{i+j}$.
Transpose the cofactor matrix to get the adjugate, $\text{adj}(A)$.
Divide every entry of the adjugate by $\det A$.
The sign pattern for the cofactors of a 3x3 is a checkerboard:
$$\begin{bmatrix} + & - & + \ - & + & - \ + & - & + \end{bmatrix}.$$
Miss one of those signs and the whole inverse is wrong — which is exactly where most errors live.
How Do You Find the Inverse of a 3x3 Matrix Using Row Operations?
The adjugate method is reliable but cofactor-heavy. The row-reduction (Gauss-Jordan) method trades the cofactors for bookkeeping. Augment $A$ with the identity to form $[A \mid I]$, then apply elementary row operations until the left block becomes the identity:
$$[,A \mid I,] ;\xrightarrow{\text{row operations}}; [,I \mid A^{-1},].$$
Whatever the right block has become is $A^{-1}$. The operations allowed are the three classics — swap two rows, scale a row by a non-zero constant, add a multiple of one row to another. If the left block can never reach the identity (a row of zeros appears), the matrix is singular and has no inverse, the same verdict the determinant gives. For a 3x3, row reduction and the adjugate formula take comparable effort; for $4 \times 4$ and larger, row reduction wins decisively, which is why it scales to bigger systems.
Examples of Inverse of 3x3 Matrix
The set builds from a quick invertibility check, through the most common sign mistake, to a full adjugate inverse, a row-reduction inverse, a diagonal special case, and a system solved by inversion.
Example 1
Is $A = \begin{bmatrix} 2 & 1 & 1 \ 1 & 3 & 2 \ 1 & 0 & 0 \end{bmatrix}$ invertible?
Compute the determinant by expanding along the bottom row (it has two zeros, so it is fastest):
$$\det A = 1 \cdot \big(1\cdot 2 - 1\cdot 3\big) - 0 + 0 = 1(2 - 3) = -1.$$
Final answer: $\det A = -1 \neq 0$, so $A$ is invertible. Expanding along the row with the most zeros saves real work.
Example 2
Find the cofactor $C_{12}$ for $A = \begin{bmatrix} 4 & 2 & 1 \ 3 & 5 & 2 \ 1 & 0 & 6 \end{bmatrix}$.
Wrong attempt. A student deletes row 1 and column 2, takes the minor $M_{12} = \det\begin{bmatrix} 3 & 2 \ 1 & 6 \end{bmatrix} = 18 - 2 = 16$, and writes $C_{12} = 16$. The minor is right; the cofactor is not.
Check the position. Entry $(1,2)$ sits on a minus square of the checkerboard, so the sign $(-1)^{1+2} = -1$ must be applied. A positive cofactor here cannot be correct.
Correct. Attach the sign:
$$C_{12} = (-1)^{1+2} M_{12} = -16.$$
Final answer: $C_{12} = -16$. The minor measures size; the cofactor carries the position's sign as well.
Example 3
Find the inverse of $A = \begin{bmatrix} 1 & 2 & 3 \ 0 & 1 & 4 \ 5 & 6 & 0 \end{bmatrix}$ using the adjugate formula.
Determinant, expanding along the first row:
$$\det A = 1(1\cdot 0 - 4\cdot 6) - 2(0\cdot 0 - 4\cdot 5) + 3(0\cdot 6 - 1\cdot 5) = -24 + 40 - 15 = 1.$$
A determinant of 1 means the inverse will have whole-number entries. Build the cofactor matrix, applying the checkerboard signs:
$$C = \begin{bmatrix} -24 & 20 & -5 \ 18 & -15 & 4 \ 5 & -4 & 1 \end{bmatrix}.$$
Transpose to get the adjugate, then divide by $\det A = 1$:
$$A^{-1} = \frac{1}{1}\begin{bmatrix} -24 & 18 & 5 \ 20 & -15 & -4 \ -5 & 4 & 1 \end{bmatrix}.$$
Final answer: the matrix above. Quick check: multiply $A A^{-1}$ and confirm the identity appears — always verify a 3x3 inverse, because one slipped sign is invisible until you test it.
Example 4
Find the inverse of $A = \begin{bmatrix} 1 & 1 & 0 \ 0 & 1 & 1 \ 0 & 0 & 1 \end{bmatrix}$ using row operations.
Augment with the identity and reduce. The matrix is already upper triangular with ones on the diagonal, so only two operations are needed:
$$\left[\begin{array}{ccc|ccc} 1 & 1 & 0 & 1 & 0 & 0 \ 0 & 1 & 1 & 0 & 1 & 0 \ 0 & 0 & 1 & 0 & 0 & 1 \end{array}\right].$$
Use $R_2 \to R_2 - R_3$, then $R_1 \to R_1 - R_2$ (with the updated $R_2$):
$$\left[\begin{array}{ccc|ccc} 1 & 0 & 0 & 1 & -1 & 1 \ 0 & 1 & 0 & 0 & 1 & -1 \ 0 & 0 & 1 & 0 & 0 & 1 \end{array}\right].$$
Final answer: $A^{-1} = \begin{bmatrix} 1 & -1 & 1 \ 0 & 1 & -1 \ 0 & 0 & 1 \end{bmatrix}$. Triangular matrices invert into triangular matrices — a structural shortcut worth recognising.
Example 5
Find the inverse of the diagonal matrix $D = \begin{bmatrix} 2 & 0 & 0 \ 0 & -3 & 0 \ 0 & 0 & 5 \end{bmatrix}$.
A diagonal matrix inverts entry by entry — replace each diagonal value with its reciprocal, leaving the zeros alone:
$$D^{-1} = \begin{bmatrix} \tfrac{1}{2} & 0 & 0 \ 0 & -\tfrac{1}{3} & 0 \ 0 & 0 & \tfrac{1}{5} \end{bmatrix}.$$
Final answer: the matrix above. It exists because no diagonal entry is zero — a diagonal matrix is singular the moment any diagonal entry is.
Example 6
Solve the system $;x + 2y + 3z = 6,; y + 4z = 5,; 5x + 6y = 7;$ using the inverse from Example 3.
In matrix form $A\mathbf{x} = \mathbf{b}$ with $A$ from Example 3 and $\mathbf{b} = (6, 5, 7)^T$. Multiply both sides on the left by $A^{-1}$:
$$\mathbf{x} = A^{-1}\mathbf{b} = \begin{bmatrix} -24 & 18 & 5 \ 20 & -15 & -4 \ -5 & 4 & 1 \end{bmatrix}\begin{bmatrix} 6 \ 5 \ 7 \end{bmatrix} = \begin{bmatrix} -144 + 90 + 35 \ 120 - 75 - 28 \ -30 + 20 + 7 \end{bmatrix} = \begin{bmatrix} -19 \ 17 \ -3 \end{bmatrix}.$$
Final answer: $x = -19,; y = 17,; z = -3$. One inverse solves the system for any right-hand side $\mathbf{b}$ — that reusability is why the inverse matters beyond a single problem.
Why the 3x3 Inverse Earns Its Place
"How do we send every point back where it came from?"
The matrix inverse was put on firm footing by Arthur Cayley (1821–1895, England), whose 1858 A Memoir on the Theory of Matrices gave the inverse, the identity, and matrix multiplication their modern algebra. Before that, "solving a system" meant grinding through elimination by hand every time.
Where the 3x3 inverse does real work today:
Solving 3D systems in one move. Three equations, three unknowns — circuit currents, chemical balances, intersecting planes — collapse to $\mathbf{x} = A^{-1}\mathbf{b}$, reusable for every new right-hand side.
Computer graphics. Undoing a 3D rotation, scale, and shear means applying the inverse transformation — every camera that "looks back" along its view does a matrix inversion underneath.
Cryptography. The Hill cipher encrypts text by multiplying blocks by a 3x3 key matrix; decryption multiplies by its inverse, which only exists when the key's determinant is invertible.
Robotics and control. Inverse kinematics — working out the joint angles that place a robot's hand at a target — leans on inverting 3x3 (and larger) matrices.
Where Students Trip Up on the Inverse of a 3x3 Matrix
Mistake 1: Forgetting to check the determinant first
Where it slips in: Diving straight into cofactors without computing $\det A$.
Don't do this: Build the entire adjugate, then divide by a determinant that turns out to be zero.
The correct way: Compute $\det A$ first. If it is zero, the matrix is singular and the inverse does not exist — no amount of cofactor work changes that.
Mistake 2: Dropping a cofactor sign
Where it slips in: Filling the cofactor matrix from the minors without applying the $(-1)^{i+j}$ checkerboard.
Don't do this: Copy each minor straight into the cofactor matrix as a positive number.
The correct way: Attach the sign for every position — the $+,-,+,/,-,+,-,/,+,-,+$ pattern. The second-guesser tends to get the signs right, then "fix" them back to positive on a recheck; trusting the pattern the first time avoids that.
Mistake 3: Transposing the wrong matrix (or forgetting to transpose)
Where it slips in: Going from the cofactor matrix to the adjugate.
Don't do this: Divide the cofactor matrix by the determinant directly — that skips the transpose and gives the wrong inverse.
The correct way: The adjugate is the transpose of the cofactor matrix. Transpose first, then divide. The rusher who skips this gets an answer that fails the $A A^{-1} = I$ check — which is why verifying is non-negotiable.
Key Takeaways
The inverse of a 3x3 matrix $A$ satisfies $AA^{-1} = I$ and exists only when $\det A \neq 0$.
The adjugate formula is $A^{-1} = \frac{1}{\det A},\text{adj}(A)$, where the adjugate is the transpose of the cofactor matrix.
Always compute the determinant first — a zero determinant means no inverse, full stop.
The cofactor checkerboard $+,-,+,/,-,+,-,/,+,-,+$ is where most sign errors creep in.
Row reduction on $[A \mid I]$ gives the same inverse and scales to larger matrices.
One inverse solves $A\mathbf{x} = \mathbf{b}$ for every right-hand side $\mathbf{b}$ — the reason it powers 3D graphics, ciphers, and robotics.
Practice These Before Moving On
Find the determinant of $\begin{bmatrix} 2 & 0 & 1 \ 3 & 1 & 2 \ 1 & 0 & 1 \end{bmatrix}$ and state whether it is invertible.
Build the cofactor matrix of $\begin{bmatrix} 1 & 0 & 2 \ 0 & 3 & 0 \ 4 & 0 & 5 \end{bmatrix}$.
Find the inverse of the diagonal matrix $\begin{bmatrix} 4 & 0 & 0 \ 0 & 2 & 0 \ 0 & 0 & -1 \end{bmatrix}$.
Answer to Question 1: $\det = 1 \neq 0$, so it is invertible. Answer to Question 2: cofactor matrix $\begin{bmatrix} 15 & 0 & -12 \ 0 & -3 & 0 \ -6 & 0 & 3 \end{bmatrix}$. Answer to Question 3: $\begin{bmatrix} \tfrac{1}{4} & 0 & 0 \ 0 & \tfrac{1}{2} & 0 \ 0 & 0 & -1 \end{bmatrix}$. If Question 2 came out all-positive, return to Mistake 2 and recheck the checkerboard signs.
Want a live Bhanzu trainer to walk through more inverse of 3x3 matrix problems? Book a free demo class — online globally.
Was this article helpful?
Your feedback helps us write better content