Matrix Equation: How to Solve AX = B with Examples

The Equation That Solves a Hundred Equations at Once

A power-grid engineer can face a hundred linear equations that must all hold at once. Economists modelling a market and graphics programmers lighting a scene hit the same wall, and writing those equations one by one is hopeless. The matrix equation collapses the entire system into three letters, $AX = B$, and hands it to a single, reliable solving rule. One compact statement, and the whole tangle becomes solvable.

What a Matrix Equation Is

A matrix equation is an equation of the form $AX = B$, where $A$ is the coefficient matrix, $X$ is the column matrix of unknown variables, and $B$ is the column matrix of constants. It is the matrix form of a system of linear equations: every equation in the system becomes one row of $A$ paired with one entry of $B$.

For example, the system

$$2x + 3y = 5$$

$$4x - y = 1$$

becomes the matrix equation

$$\begin{bmatrix} 2 & 3 \ 4 & -1 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 5 \ 1 \end{bmatrix}$$

The first matrix is $A$, the column of variables is $X$, and the column of constants is $B$. Solving the matrix equation means finding the $X$ that makes the product equal $B$. This is the same problem as solving a system of equations — only written so a single operation can finish it.

How Do You Write a System as a Matrix Equation?

Three steps turn any linear system into $AX = B$.

1. Line up the variables. Write each equation with the same variables in the same order, filling in a coefficient of 0 where a variable is missing.

2. Build $A$ from the coefficients. Each row of $A$ is one equation's coefficients, in variable order.

3. Build $X$ and $B$. $X$ is the column of variables; $B$ is the column of the right-hand-side constants.

Take the system $3x - 2y = 7$ and $x + 5y = -4$. Lined up, the coefficients are $(3, -2)$ and $(1, 5)$, so

$$A = \begin{bmatrix} 3 & -2 \ 1 & 5 \end{bmatrix}, \quad X = \begin{bmatrix} x \ y \end{bmatrix}, \quad B = \begin{bmatrix} 7 \ -4 \end{bmatrix}$$

The missing-variable rule matters. If one equation reads $2x = 6$ in a two-variable system, its row is $(2, 0)$, not $(2)$ — the zero holds the place of $y$.

How to Solve a Matrix Equation

The standard route is the inverse matrix method. Just as you solve $3x = 12$ by multiplying both sides by $\tfrac{1}{3}$, you solve $AX = B$ by multiplying both sides on the left by $A^{-1}$:

$$AX = B$$

$$A^{-1}AX = A^{-1}B$$

$$IX = A^{-1}B$$

$$X = A^{-1}B$$

Here $I$ is the identity matrix, which leaves $X$ unchanged, so the solution is

$$\boxed{X = A^{-1}B}$$

The whole method rests on $A$ having an inverse. The inverse of a matrix exists only when its determinant is non-zero. So before solving, check $\det A$.

Variable glossary. $A$ is the coefficient matrix; $X$ is the column of unknowns; $B$ is the column of constants; $A^{-1}$ is the inverse of $A$; $I$ is the identity matrix; $\det A$ is the determinant of $A$.

A note on which method to reach for. The inverse method is clean for a $2 \times 2$ system you are solving once. For larger systems, finding $A^{-1}$ is heavy, and Cramer's rule or Gaussian elimination is usually faster and more numerically stable. The inverse method teaches the idea best; the others scale better.

When Does a Matrix Equation Have a Solution?

The determinant decides everything.

• $\det A \neq 0$. The inverse exists, and the equation has exactly one solution, $X = A^{-1}B$.

• $\det A = 0$ (singular $A$). The inverse does not exist, and the system has either no solution or infinitely many — never exactly one. Which case you are in depends on whether $B$ is consistent with the rows of $A$.

This is why the determinant check is not optional bookkeeping. It is the test for whether the inverse method can even run.

Examples of Matrix Equation

Example 1

Write the system $x + 2y = 4$, $3x - y = 5$ as a matrix equation.

$$\begin{bmatrix} 1 & 2 \ 3 & -1 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 4 \ 5 \end{bmatrix}$$

Final answer: $A = \begin{bmatrix} 1 & 2 \ 3 & -1 \end{bmatrix}$, $X = \begin{bmatrix} x \ y \end{bmatrix}$, $B = \begin{bmatrix} 4 \ 5 \end{bmatrix}$.

Example 2

Solve $\begin{bmatrix} 2 & 1 \ 1 & 3 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 5 \ 10 \end{bmatrix}$.

First, $\det A = (2)(3) - (1)(1) = 5 \neq 0$, so the inverse exists. For a $2 \times 2$ matrix, the inverse is $\tfrac{1}{\det A}\begin{bmatrix} d & -b \ -c & a \end{bmatrix}$:

$$A^{-1} = \frac{1}{5}\begin{bmatrix} 3 & -1 \ -1 & 2 \end{bmatrix}$$

Now $X = A^{-1}B$:

$$X = \frac{1}{5}\begin{bmatrix} 3 & -1 \ -1 & 2 \end{bmatrix}\begin{bmatrix} 5 \ 10 \end{bmatrix} = \frac{1}{5}\begin{bmatrix} 15 - 10 \ -5 + 20 \end{bmatrix} = \frac{1}{5}\begin{bmatrix} 5 \ 15 \end{bmatrix} = \begin{bmatrix} 1 \ 3 \end{bmatrix}$$

Final answer: $x = 1$, $y = 3$.

Example 3 (watch how this goes wrong)

Solve $AX = B$ where $A = \begin{bmatrix} 1 & 2 \ 2 & 4 \end{bmatrix}$ and $B = \begin{bmatrix} 3 \ 7 \end{bmatrix}$.

Wrong attempt. A student computes $A^{-1}$ using the $2 \times 2$ formula and pushes straight to $X = A^{-1}B$, expecting a clean answer.

The break. Check the determinant first: $\det A = (1)(4) - (2)(2) = 4 - 4 = 0$. The matrix is singular, so $A^{-1}$ does not exist — the $\tfrac{1}{\det A}$ factor would divide by zero. The inverse method cannot run here.

Correct. Read the two equations: $x + 2y = 3$ and $2x + 4y = 7$. The second is twice the first on the left ($2x + 4y$), but $2 \times 3 = 6 \neq 7$ on the right. The lines are parallel and never meet. Final answer: no solution (the system is inconsistent).

Example 4

A right-hand side that flips the verdict: solve $\begin{bmatrix} 1 & 2 \ 2 & 4 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 3 \ 6 \end{bmatrix}$.

Same singular $A$ as Example 3, but now $B = \begin{bmatrix} 3 \ 6 \end{bmatrix}$. The equations are $x + 2y = 3$ and $2x + 4y = 6$ — the second is exactly twice the first, so they are the same line.

Final answer: infinitely many solutions, all points on $x + 2y = 3$. (Compare with Example 3: same $A$, different $B$, opposite outcome.)

Example 5

Solve the $3 \times 3$ system $x + y + z = 6$, $2y + 5z = -4$, $2x + 5y - z = 27$ using the matrix equation, given $\det A = -21$.

$$A = \begin{bmatrix} 1 & 1 & 1 \ 0 & 2 & 5 \ 2 & 5 & -1 \end{bmatrix}, \quad B = \begin{bmatrix} 6 \ -4 \ 27 \end{bmatrix}$$

Since $\det A = -21 \neq 0$, a unique solution exists. Solving $X = A^{-1}B$ (or by elimination) gives

$$x = 5, \quad y = 3, \quad z = -2$$

Check in equation 1: $5 + 3 + (-2) = 6$ ✓. Final answer: $(x, y, z) = (5, 3, -2)$.

Example 6

Find the unknown matrix $X$ in $\begin{bmatrix} 1 & 0 \ 0 & 2 \end{bmatrix}X = \begin{bmatrix} 4 \ 6 \end{bmatrix}$.

Here $A$ is diagonal, so $A^{-1} = \begin{bmatrix} 1 & 0 \ 0 & \tfrac{1}{2} \end{bmatrix}$:

$$X = \begin{bmatrix} 1 & 0 \ 0 & \frac{1}{2} \end{bmatrix}\begin{bmatrix} 4 \ 6 \end{bmatrix} = \begin{bmatrix} 4 \ 3 \end{bmatrix}$$

Final answer: $X = \begin{bmatrix} 4 \ 3 \end{bmatrix}$, i.e. $x = 4$, $y = 3$.

Why Matrix Equations Matter — "one rule for any system"

Long before computers, people had systems of equations to solve in surveying, astronomy, and trade, and they solved them by repetitive substitution. The breakthrough was realising that the structure of every linear system is the same — coefficients times unknowns equals constants — so a single notation and a single method could handle all of them. That is what $AX = B$ delivers.

• Engineering. Circuit analysis, structural load balancing, and control systems all reduce to large $AX = B$ systems solved by computer.

• Computer graphics and physics simulation. Lighting, fluid flow, and rigid-body dynamics are systems of linear equations solved thousands of times per second.

• Economics and data science. Linear regression — fitting a line or surface to data — is solved as a matrix equation, and so are equilibrium and optimisation models.

A detail worth carrying: the reason software almost never literally computes $A^{-1}$ for large systems is that inverting is both expensive and error-prone. The matrix equation tells the computer what to solve; methods like LU decomposition tell it how to do so cheaply. The equation is the idea; the algorithm is the engineering.

Common Errors When Working With Matrix Equations

Mistake 1: Solving without checking the determinant

Where it slips in: Jumping straight to $X = A^{-1}B$.

Don't do this: Computing the inverse before confirming $A$ is invertible. If $\det A = 0$, the inverse formula divides by zero and the method silently breaks.

The correct way: Compute $\det A$ first. If it is non-zero, proceed; if it is zero, the system has no solution or infinitely many, and you analyse the equations directly. The first instinct is to treat every system as if it has a clean unique answer, and the determinant is the one check that catches the systems that do not.

Mistake 2: Multiplying on the wrong side

Where it slips in: Applying $A^{-1}$ to solve $AX = B$.

Don't do this: Writing $X = BA^{-1}$. Matrix multiplication is not commutative, so the side matters.

The correct way: Multiply $A^{-1}$ on the left of both sides, since $A$ sits to the left of $X$. That gives $X = A^{-1}B$. The habit that fixes this is writing the inverse on the same side as the matrix you are cancelling, every time.

Mistake 3: Dropping a zero coefficient when building A

Where it slips in: A system where one equation is missing a variable.

Don't do this: Writing a short row that skips the missing variable, which misaligns the columns of $A$.

The correct way: Put a 0 in the slot for any missing variable so every row has the same length and the columns stay aligned. A misaligned $A$ produces a confident, wrong answer.

Bottom Line

• A matrix equation writes a linear system compactly as $AX = B$, with $A$ the coefficients, $X$ the unknowns, and $B$ the constants.

• When $A$ is invertible, the solution is $X = A^{-1}B$.

• A unique solution exists exactly when $\det A \neq 0$; a singular $A$ means no solution or infinitely many.

• The most common mistake is solving without first checking the determinant.

• For large systems, elimination or Cramer's rule beats computing the inverse directly.

Practice Questions on Matrix Equation

Work through these, then check your answers below.

1. Rewrite the system $4x - y = 3$, $2x + 3y = 11$ as a matrix equation $AX = B$.

2. Solve the system from Question 1 by the inverse method.

3. Without fully solving, decide whether $\begin{bmatrix} 1 & 3 \ 2 & 6 \end{bmatrix}X = \begin{bmatrix} 2 \ 5 \end{bmatrix}$ has a solution.

4. Solve $\begin{bmatrix} 1 & 1 \ 1 & -1 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 6 \ 2 \end{bmatrix}$.

Answer to Question 1: $\begin{bmatrix} 4 & -1 \ 2 & 3 \end{bmatrix}\begin{bmatrix} x \ y \end{bmatrix} = \begin{bmatrix} 3 \ 11 \end{bmatrix}$.

Answer to Question 2: $\det A = (4)(3) - (-1)(2) = 14 \neq 0$, so $A^{-1} = \tfrac{1}{14}\begin{bmatrix} 3 & 1 \ -2 & 4 \end{bmatrix}$. Then $X = \tfrac{1}{14}\begin{bmatrix} 3 & 1 \ -2 & 4 \end{bmatrix}\begin{bmatrix} 3 \ 11 \end{bmatrix} = \tfrac{1}{14}\begin{bmatrix} 20 \ 38 \end{bmatrix} = \begin{bmatrix} \frac{10}{7} \ \frac{19}{7} \end{bmatrix}$. So $x = \tfrac{10}{7}$, $y = \tfrac{19}{7}$.

Answer to Question 3: $\det A = (1)(6) - (3)(2) = 0$, so $A$ is singular. Checking consistency: the first equation is $x + 3y = 2$ and the second is $2x + 6y = 5$. Doubling the first gives $2x + 6y = 4 \neq 5$, so the system is inconsistent and has no solution.

Answer to Question 4: $\det A = (1)(-1) - (1)(1) = -2 \neq 0$, so $A^{-1} = \tfrac{1}{-2}\begin{bmatrix} -1 & -1 \ -1 & 1 \end{bmatrix} = \tfrac{1}{2}\begin{bmatrix} 1 & 1 \ 1 & -1 \end{bmatrix}$. Then $X = \tfrac{1}{2}\begin{bmatrix} 1 & 1 \ 1 & -1 \end{bmatrix}\begin{bmatrix} 6 \ 2 \end{bmatrix} = \tfrac{1}{2}\begin{bmatrix} 8 \ 4 \end{bmatrix} = \begin{bmatrix} 4 \ 2 \end{bmatrix}$. So $x = 4$, $y = 2$.

A Practical Next Step

If the determinant step trips you, return to the "When Does a Matrix Equation Have a Solution?" section. At Bhanzu, our trainers teach students to compute the determinant before reaching for the inverse, so a singular system is caught at the start rather than after a page of wrong work.

Want a live Bhanzu trainer to walk through more matrix equation problems? Book a free demo class.