diff --git a/libavcodec/avcodec.h b/libavcodec/avcodec.h
index c2f768723476eabdc48b920667c1af50073b5ba8..dfd7264458f31ef6d813672db7c9902d3597ba31 100644
--- a/libavcodec/avcodec.h
+++ b/libavcodec/avcodec.h
@@ -2104,7 +2104,7 @@ typedef struct AVCodecContext {
 #define X264_PART_B8X8 0x100  /* Analyze b16x8, b8x16 and b8x8 */
 
     /**
-     * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal)
+     * direct MV prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)
      * - encoding: Set by user.
      * - decoding: unused
      */
diff --git a/libavcodec/utils.c b/libavcodec/utils.c
index cf00d2552cf6edcc84618f582a9bfca630dbd7ee..e56b9870a56b24a79ef3809ec9e6c0cde648e78d 100644
--- a/libavcodec/utils.c
+++ b/libavcodec/utils.c
@@ -691,7 +691,7 @@ static const AVOption options[]={
 {"chromaoffset", "chroma qp offset from luma", OFFSET(chromaoffset), FF_OPT_TYPE_INT, DEFAULT, INT_MIN, INT_MAX, V|E},
 {"bframebias", "influences how often B-frames are used", OFFSET(bframebias), FF_OPT_TYPE_INT, DEFAULT, INT_MIN, INT_MAX, V|E},
 {"trellis", "rate-distortion optimal quantization", OFFSET(trellis), FF_OPT_TYPE_INT, DEFAULT, INT_MIN, INT_MAX, V|A|E},
-{"directpred", "direct mv prediction mode - 0 (none), 1 (spatial), 2 (temporal)", OFFSET(directpred), FF_OPT_TYPE_INT, 2, INT_MIN, INT_MAX, V|E},
+{"directpred", "direct mv prediction mode - 0 (none), 1 (spatial), 2 (temporal), 3 (auto)", OFFSET(directpred), FF_OPT_TYPE_INT, 2, INT_MIN, INT_MAX, V|E},
 {"bpyramid", "allows B-frames to be used as references for predicting", 0, FF_OPT_TYPE_CONST, CODEC_FLAG2_BPYRAMID, INT_MIN, INT_MAX, V|E, "flags2"},
 {"wpred", "weighted biprediction for b-frames (H.264)", 0, FF_OPT_TYPE_CONST, CODEC_FLAG2_WPRED, INT_MIN, INT_MAX, V|E, "flags2"},
 {"mixed_refs", "one reference per partition, as opposed to one reference per macroblock", 0, FF_OPT_TYPE_CONST, CODEC_FLAG2_MIXED_REFS, INT_MIN, INT_MAX, V|E, "flags2"},