What Is The Transpose Of A Matrix?
The transpose of a matrix $A$ is a new matrix, written $A^T$ (or sometimes $A'$), formed by interchanging the rows and columns of $A$. The entry that sits in row $i$ and column $j$ of $A$ lands in row $j$ and column $i$ of $A^T$.
In symbols, if $A = [a_{ij}]$, then $A^T = [a_{ji}]$.
Variable glossary. $a_{ij}$ is the entry in row $i$, column $j$ of $A$. The superscript $T$ marks the transpose. The order of a matrix is its row-count by column-count, written $m \times n$.
Take a small example:
$$A = \begin{bmatrix} 2 & 5 & 1 \ 7 & 9 & 4 \end{bmatrix}, \qquad A^T = \begin{bmatrix} 2 & 7 \ 5 & 9 \ 1 & 4 \end{bmatrix}$$
The first row of $A$, namely $2, 5, 1$, becomes the first column of $A^T$. The second row becomes the second column. Nothing is added or removed; the numbers are simply read in the other direction.
How Does The Order Change When You Transpose?
If $A$ has order $m \times n$, then $A^T$ has order $n \times m$. The dimensions swap.
A $2 \times 3$ matrix transposes to a $3 \times 2$ matrix. A $4 \times 1$ column vector transposes to a $1 \times 4$ row vector. Only a square matrix keeps the same order after transposing, because $n \times n$ stays $n \times n$.
This order-swap is the detail that controls everything later. It is exactly why the transpose of a product reverses — the orders have to stay compatible for multiplication of matrices to remain defined.
What Are The Properties Of The Transpose?
Five properties carry almost every transpose problem you will meet. Each one has a short reason behind it, not just a rule to memorise.
Transpose of a transpose returns the original: $(A^T)^T = A$. Swap rows and columns, then swap them back, and you are where you started.
Transpose distributes over addition: $(A + B)^T = A^T + B^T$. Adding happens entry by entry, and transposing only moves entries to new addresses, so the two operations do not interfere.
Scalars pass straight through: $(kA)^T = k,A^T$ for any scalar $k$. Multiplying every entry by $k$ commutes with relabelling positions. (This is the same scalar that drives matrix scalar multiplication.)
Transpose of a product reverses the order: $(AB)^T = B^T A^T$. This is the rule worth slowing down on — see the worked examples.
Determinant is unchanged: $\det(A^T) = \det(A)$ for a square matrix. The determinant reads the same whether you expand along rows or columns.
The reversal rule generalises to longer products: $(ABC)^T = C^T B^T A^T$. The factors come back in reverse order, each one transposed.
How Does The Transpose Define Symmetric Matrices?
The transpose is what makes "symmetric" a precise idea instead of a vague one.
A square matrix is symmetric when it equals its own transpose: $A = A^T$. The entries mirror across the main diagonal — $a_{ij} = a_{ji}$ for every pair. A square matrix is skew-symmetric when $A^T = -A$, which forces every diagonal entry to be 0.
$$\begin{bmatrix} 1 & 4 & 7 \ 4 & 2 & 5 \ 7 & 5 & 3 \end{bmatrix} \text{ is symmetric, since it equals its transpose.}$$
Examples Of The Transpose Of A Matrix
Example 1
Find the transpose of $A = \begin{bmatrix} 3 & 8 \ 1 & 6 \end{bmatrix}$.
Read each row of $A$ as a column.
$$A^T = \begin{bmatrix} 3 & 1 \ 8 & 6 \end{bmatrix}$$
The diagonal entries 3 and 6 stay put; the off-diagonal pair 8 and 1 swap places.
Final answer: $A^T = \begin{bmatrix} 3 & 1 \ 8 & 6 \end{bmatrix}$.
Example 2
Transpose the rectangular matrix $B = \begin{bmatrix} 2 & -1 & 0 \ 4 & 5 & 3 \end{bmatrix}$.
$B$ is $2 \times 3$, so $B^T$ is $3 \times 2$. The first row becomes the first column:
$$B^T = \begin{bmatrix} 2 & 4 \ -1 & 5 \ 0 & 3 \end{bmatrix}$$
Final answer: $B^T = \begin{bmatrix} 2 & 4 \ -1 & 5 \ 0 & 3 \end{bmatrix}$.
Example 3
Find $(AB)^T$ for $A = \begin{bmatrix} 1 & 2 \ 0 & 1 \end{bmatrix}$ and $B = \begin{bmatrix} 3 & 0 \ 4 & 5 \end{bmatrix}$.
The first instinct is to compute $A^T B^T$, transposing each factor and keeping the order.
Wrong attempt.
$A^T = \begin{bmatrix} 1 & 0 \ 2 & 1 \end{bmatrix}$, $B^T = \begin{bmatrix} 3 & 4 \ 0 & 5 \end{bmatrix}$
$A^T B^T = \begin{bmatrix} 3 & 4 \ 6 & 13 \end{bmatrix}$
Why it is wrong. Compute $AB$ first, then transpose it, and the answers disagree.
$AB = \begin{bmatrix} 11 & 10 \ 4 & 5 \end{bmatrix}$
$(AB)^T = \begin{bmatrix} 11 & 4 \ 10 & 5 \end{bmatrix}$
That is not $\begin{bmatrix} 3 & 4 \ 6 & 13 \end{bmatrix}$, so $A^T B^T$ cannot be right.
Correct method. The rule is $(AB)^T = B^T A^T$ — reverse the order.
$B^T A^T = \begin{bmatrix} 3 & 4 \ 0 & 5 \end{bmatrix}\begin{bmatrix} 1 & 0 \ 2 & 1 \end{bmatrix} = \begin{bmatrix} 11 & 4 \ 10 & 5 \end{bmatrix}$
This matches $(AB)^T$ exactly.
Final answer: $(AB)^T = B^T A^T = \begin{bmatrix} 11 & 4 \ 10 & 5 \end{bmatrix}$.
Example 4
Verify $(A^T)^T = A$ for $A = \begin{bmatrix} 5 & -2 & 6 \end{bmatrix}$.
$A$ is a $1 \times 3$ row vector. Its transpose is a $3 \times 1$ column:
$A^T = \begin{bmatrix} 5 \ -2 \ 6 \end{bmatrix}$
Transpose again, and the column folds back into the original row:
$(A^T)^T = \begin{bmatrix} 5 & -2 & 6 \end{bmatrix} = A$
Final answer: $(A^T)^T = A$, confirmed.
Example 5
Is $A = \begin{bmatrix} 2 & 9 \ 9 & 7 \end{bmatrix}$ symmetric?
Find the transpose and compare.
$A^T = \begin{bmatrix} 2 & 9 \ 9 & 7 \end{bmatrix}$
Because $A^T$ equals $A$, the matrix is symmetric. The off-diagonal entries match: $a_{12} = a_{21} = 9$.
Final answer: Yes, $A$ is symmetric, since $A = A^T$.
Example 6
Confirm the scalar rule $(3A)^T = 3A^T$ for $A = \begin{bmatrix} 1 & 4 \ 0 & 2 \end{bmatrix}$.
Left side: multiply first, then transpose.
$3A = \begin{bmatrix} 3 & 12 \ 0 & 6 \end{bmatrix}$
$(3A)^T = \begin{bmatrix} 3 & 0 \ 12 & 6 \end{bmatrix}$
Right side: transpose first, then multiply.
$A^T = \begin{bmatrix} 1 & 0 \ 4 & 2 \end{bmatrix}$
$3A^T = \begin{bmatrix} 3 & 0 \ 12 & 6 \end{bmatrix}$
Both sides agree.
Final answer: $(3A)^T = 3A^T = \begin{bmatrix} 3 & 0 \ 12 & 6 \end{bmatrix}$.
Why The Transpose Matters: "The bridge between rows and columns"
The transpose exists because rows and columns are not interchangeable inside a matrix, yet many real computations need to treat them as if they were. The transpose is the formal bridge that lets you cross between the two.
Where that bridge does real work:
Least-squares and regression. Fitting a line or model to data leans on the normal equations, $A^T A,x = A^T b$. The transpose appears on both sides; without it, there is no clean way to project data onto a model.
Symmetry detection. Covariance matrices in statistics are symmetric by construction because they are built as $A^T A$. The transpose guarantees the symmetry that downstream methods rely on.
How computers store grids. An image, a spreadsheet, or a tensor can be laid out row-major or column-major in memory. Transposing converts between the two — a routine operation in graphics and machine-learning libraries.
A reasoning-first approach helps here because students who memorise "$(AB)^T = B^T A^T$" as a string of symbols cannot reconstruct it under pressure.
Tie it to the order rule instead: $A$ is $m \times n$ and $B$ is $n \times p$, so $AB$ is $m \times p$ and $(AB)^T$ is $p \times m$ — the only way to build a $p \times m$ product from the transposes is $B^T A^T$. At Bhanzu, trainers teach the reversal through that dimension check, so the rule has a reason a student can rebuild rather than a phrase they hope to recall. Explore the approach.
Where Students Slip On The Transpose (And How To Fix It)
Mistake 1: Forgetting to reverse the order in a product
Where it slips in: Any time a transpose lands on $AB$.
Don't do this: Write $(AB)^T = A^T B^T$, keeping the factors in their original order.
The correct way: Reverse them: $(AB)^T = B^T A^T$. The first instinct is to transpose each factor in place and leave the order alone — that is the single most common transpose error, and it usually surfaces only when the orders are rectangular and the wrong version does not even form a valid product. Run the dimension check every time.
Mistake 2: Transposing a non-square matrix without changing its order
Where it slips in: Transposing a rectangular matrix in a hurry.
Don't do this: Keep a $2 \times 3$ matrix at $2 \times 3$ after transposing, just rearranging entries inside the same shape.
The correct way: A $2 \times 3$ matrix becomes $3 \times 2$. The rusher who treats transposing as "shuffle the numbers" loses the order swap; the memorizer who only ever practised on square matrices never had to confront it. Write the new dimensions before filling in entries.
Mistake 3: Confusing the transpose with the inverse
Where it slips in: Reading $A^T$ as "the matrix that undoes $A$."
Don't do this: Assume $A A^T = I$ for every matrix.
The correct way: $A^T$ is the inverse only for the special orthogonal-matrix family, where $A^T A = I$ holds. For a general matrix, the transpose and the inverse are different objects entirely: the transpose always exists, the inverse may not.
Key Takeaways
The transpose of a matrix swaps its rows and columns: the entry $a_{ij}$ moves to position $(j, i)$, and $A$ of order $m \times n$ becomes $A^T$ of order $n \times m$.
The core properties are $(A^T)^T = A$, $(A + B)^T = A^T + B^T$, $(kA)^T = k A^T$, and $\det(A^T) = \det(A)$.
The transpose of a product reverses the factor order: $(AB)^T = B^T A^T$.
A matrix is symmetric when $A = A^T$ and skew-symmetric when $A^T = -A$.
The transpose underpins least-squares regression, covariance matrices, and how grids are stored in memory.
A Practical Next Step
Work through these to lock in the operation, then check the order-swap section if the product rule keeps tripping you:
Transpose $\begin{bmatrix} 1 & 2 & 3 \ 4 & 5 & 6 \end{bmatrix}$ and state its new order.
Check whether $\begin{bmatrix} 0 & -3 \ 3 & 0 \end{bmatrix}$ is skew-symmetric.
Verify $(AB)^T = B^T A^T$ for two $2 \times 2$ matrices of your choice.
To build matrix operations with a teacher who explains the why behind each rule, work with an algebra tutor, get help with algebra, or join math classes online. Want a live Bhanzu trainer to walk through more transpose problems? Book a free demo class.
Read More
Matrices and determinants — the broader picture of how matrices and their determinants fit together.
Adjoint of a matrix — the cofactor-transpose construction that builds on this operation.
Cofactor matrix — transposed, it gives the adjoint used in matrix inversion.
Solve matrices — where the transpose appears in the normal equations for solving systems.
Was this article helpful?
Your feedback helps us write better content