diff --git a/docs/user-guide/sinks/retry-strategy.md b/docs/user-guide/sinks/retry-strategy.md
new file mode 100644
index 0000000000..a5b2a7264b
--- /dev/null
+++ b/docs/user-guide/sinks/retry-strategy.md
@@ -0,0 +1,68 @@
+# Retry Strategy
+
+### Overview
+The `RetryStrategy` is used to configure the behavior for a sink after encountering failures during a write operation. 
+This structure allows the user to specify how Numaflow should respond to different fail-over scenarios for Sinks, ensuring that the writing can be resilient and handle 
+unexpected issues efficiently.
+
+
+### Struct Explanation
+
+
+`retryStrategy` is optional, and can be added to the Sink spec configurations where retry logic is necessary.
+
+
+
+```yaml
+sink:
+  retryStrategy:
+      # Optional
+      backoff:
+        duration: 1s # Optional
+        steps: 3 # Optional, number of retries (including the 1st try)
+      # Optional
+      onFailure: retry|fallback|drop 
+```
+Note: If no custom fields are defined for retryStrategy then the **default** values are used.
+
+- `BackOff` - Defines the timing for retries, including the interval and the maximum attempts.
+  - `duration`: the time interval to wait before retry attempts
+    - Default: _1ms_
+  - `steps`: the limit on the number of times to try the sink write operation including retries
+    - Default: _Infinite_
+- `OnFailure` - Specifies the action to be undertaken if number of retries are exhausted
+  - retry: continue with the retry logic again
+  - fallback: write the leftover messages to a [fallback](https://numaflow.numaproj.io/user-guide/sinks/fallback/) sink
+  - drop: any messages left to be processed are dropped
+    - Default: _retry_
+
+
+### Constraints
+
+1) If the `onFailure` is defined as fallback, then there should be a fallback sink specified in the spec.
+
+2) The steps defined should always be `> 0`
+
+
+## Example
+
+```yaml
+  sink:
+    retryStrategy:
+      backoff:
+        interval: "500ms"
+        steps: 10
+      onFailure: "fallback"
+    udsink:
+      container:
+        image: my-sink-image
+    fallback:
+      udsink:
+        container:
+          image: my-fallback-sink
+```
+### Explanation
+
+- Normal Operation: Data is processed by the primary sink container specified by `UDSink`. 
+The system retries up to 10 times for a batch write operation to succeed with an interval of 500 milliseconds between each retry.
+- After Maximum Retries: If all retries fail, data is then routed to a fallback sink instead.
diff --git a/mkdocs.yml b/mkdocs.yml
index 7c9161d480..7e57a6af21 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -66,6 +66,7 @@ nav:
           - user-guide/sinks/blackhole.md
           - User-defined Sinks: "user-guide/sinks/user-defined-sinks.md"
           - Fallback Sink: "user-guide/sinks/fallback.md"
+          - Retry Strategy: "user-guide/sinks/retry-strategy.md"
       - User-defined Functions:
           - Overview: "user-guide/user-defined-functions/user-defined-functions.md"
           - Map: