diff --git a/include/cglm-mat3-simd-sse2.h b/include/cglm-mat3-simd-sse2.h
index 4c6e7f4..aed2d71 100644
--- a/include/cglm-mat3-simd-sse2.h
+++ b/include/cglm-mat3-simd-sse2.h
@@ -11,6 +11,66 @@
 
 #include "cglm-intrin.h"
 
+CGLM_INLINE
+void
+glm_mat3_transp_to_sse2(mat3 m, mat3 dest){
+  __m128 x0, x1, x2, x3, x4;
+  /*
+   a b c d        a d g b
+   e f g h   ->   e h c f
+   j              j
+   */
+
+  /* d c b a */
+  /* h g f e */
+  x0 = _mm_loadu_ps(&m[0][0]);
+  x1 = _mm_loadu_ps(&m[1][1]);
+
+  /* g g b b */
+  /* a d g b */
+  x2 = _mm_shuffle_ps(x0, x1, _MM_SHUFFLE(2, 2, 1, 1));
+  x3 = _mm_shuffle_ps(x0, x2, _MM_SHUFFLE(0, 2, 3, 0));
+
+  /* c c f f */
+  /* e h c f */
+  x2 = _mm_shuffle_ps(x1, x0, _MM_SHUFFLE(2, 2, 1, 1));
+  x4 = _mm_shuffle_ps(x1, x2, _MM_SHUFFLE(0, 2, 3, 0));
+
+  _mm_storeu_ps(&dest[0][0], x3);
+  _mm_storeu_ps(&dest[1][1], x4);
+
+  dest[2][2] = m[2][2];
+}
+
+CGLM_INLINE
+void
+glm_mat3_transp_sse2(mat3 m){
+  __m128 x0, x1, x2, x3, x4;
+  /*
+   a b c d        a d g b
+   e f g h   ->   e h c f
+   j              j
+   */
+
+  /* d c b a */
+  /* h g f e */
+  x0 = _mm_loadu_ps(&m[0][0]);
+  x1 = _mm_loadu_ps(&m[1][1]);
+
+  /* g g b b */
+  /* a d g b */
+  x2 = _mm_shuffle_ps(x0, x1, _MM_SHUFFLE(2, 2, 1, 1));
+  x3 = _mm_shuffle_ps(x0, x2, _MM_SHUFFLE(0, 2, 3, 0));
+
+  /* c c f f */
+  /* e h c f */
+  x2 = _mm_shuffle_ps(x1, x0, _MM_SHUFFLE(2, 2, 1, 1));
+  x4 = _mm_shuffle_ps(x1, x2, _MM_SHUFFLE(0, 2, 3, 0));
+
+  _mm_storeu_ps(&m[0][0], x3);
+  _mm_storeu_ps(&m[1][1], x4);
+}
+
 CGLM_INLINE
 void
 glm_mat3_mul_sse2(mat3 m1, mat3 m2, mat3 dest) {
diff --git a/include/cglm-mat3.h b/include/cglm-mat3.h
index f17f31a..6016d8a 100644
--- a/include/cglm-mat3.h
+++ b/include/cglm-mat3.h
@@ -76,6 +76,63 @@ glm_mat3_mul(mat3 m1, mat3 m2, mat3 dest) {
 #endif
 }
 
+/*!
+ * @brief transpose mat3 and store in dest
+ *
+ * source matrix will not be transposed unless dest is m
+ *
+ * @param m[in]     matrix
+ * @param dest[out] result
+ */
+CGLM_INLINE
+void
+glm_mat3_transpose_to(mat3 m, mat3 dest) {
+#if defined( __SSE__ ) || defined( __SSE2__ )
+  glm_mat3_transp_to_sse2(m, dest);
+#else
+  dest[0][0] = m[0][0];
+  dest[0][1] = m[1][0];
+  dest[0][2] = m[2][0];
+
+  dest[1][0] = m[0][1];
+  dest[1][1] = m[1][1];
+  dest[1][2] = m[2][1];
+
+  dest[2][0] = m[0][2];
+  dest[2][1] = m[1][2];
+  dest[2][2] = m[2][2];
+#endif
+}
+
+/*!
+ * @brief tranpose mat3 and store result in same matrix
+ *
+ * @param[in, out] m source and dest
+ */
+CGLM_INLINE
+void
+glm_mat3_transpose(mat3 m) {
+#if defined( __SSE__ ) || defined( __SSE2__ )
+  glm_mat3_transp_sse2(m);
+#else
+  mat3 tmp;
+
+  tmp[0][1] = m[1][0];
+  tmp[0][2] = m[2][0];
+  tmp[1][0] = m[0][1];
+  tmp[1][2] = m[2][1];
+  tmp[2][0] = m[0][2];
+  tmp[2][1] = m[1][2];
+
+  m[0][1] = tmp[0][1];
+  m[0][2] = tmp[0][2];
+  m[1][0] = tmp[1][0];
+  m[1][2] = tmp[1][2];
+  m[2][0] = tmp[2][0];
+  m[2][1] = tmp[2][1];
+#endif
+}
+
 CGLM_INLINE
 void
 glm_mat3_print(mat3 matrix,