diff --git a/CHANGELOG.md b/CHANGELOG.md
index 83be5f478..25f9df635 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -127,6 +127,8 @@ Remember to align the itemized text with the first line of an item within a list
 
 * Changes
 
+  * JAX now supports CUDA 12.3 and CUDA 11.8. Support for CUDA 12.2 has been
+    dropped.
   * `cost_analysis` now works with cross-compiled `Compiled` objects (i.e. when
     using `.lower().compile()` with a topology object, e.g., to compile for
     Cloud TPU from a non-TPU computer).
diff --git a/docs/installation.md b/docs/installation.md
index 5dba6757e..a6fd97948 100644
--- a/docs/installation.md
+++ b/docs/installation.md
@@ -114,7 +114,7 @@ able to use the
 that NVIDIA provides for this purpose.
 
 JAX currently ships two CUDA wheel variants:
-* CUDA 12.2, cuDNN 8.9, NCCL 2.16
+* CUDA 12.3, cuDNN 8.9, NCCL 2.16
 * CUDA 11.8, cuDNN 8.6, NCCL 2.16
 
 You may use a JAX wheel provided the major version of your CUDA, cuDNN, and NCCL