diff --git a/programs/bench.c b/programs/bench.c
index 4059072f..6009ebc7 100644
--- a/programs/bench.c
+++ b/programs/bench.c
@@ -96,6 +96,7 @@ static U32 g_nbSeconds = NBSECONDS;
 static size_t g_blockSize = 0;
 static int g_additionalParam = 0;
 static U32 g_decodeOnly = 0;
+static U32 g_nbThreads = 1;
 
 void BMK_setNotificationLevel(unsigned level) { g_displayLevel=level; }
 
@@ -113,7 +114,9 @@ void BMK_SetBlockSize(size_t blockSize)
     DISPLAYLEVEL(2, "using blocks of size %u KB \n", (U32)(blockSize>>10));
 }
 
-void BMK_setDecodeOnly(unsigned decodeFlag) { g_decodeOnly = (decodeFlag>0); }
+void BMK_setDecodeOnlyMode(unsigned decodeFlag) { g_decodeOnly = (decodeFlag>0); }
+
+void BMK_SetNbThreads(unsigned nbThreads) { g_nbThreads = nbThreads; }
 
 
 /* ********************************************************
@@ -156,7 +159,7 @@ static int BMK_benchMem(const void* srcBuffer, size_t srcSize,
     U32 nbBlocks;
     UTIL_time_t ticksPerSecond;
 
-    ZSTDMT_CCtx* const mtcctx = ZSTDMT_createCCtx(1);
+    ZSTDMT_CCtx* const mtcctx = ZSTDMT_createCCtx(g_nbThreads);
 
     /* checks */
     if (!compressedBuffer || !resultBuffer || !blockTable || !ctx || !dctx)
diff --git a/programs/bench.h b/programs/bench.h
index 314f3465..87850bcc 100644
--- a/programs/bench.h
+++ b/programs/bench.h
@@ -15,14 +15,15 @@
 #define ZSTD_STATIC_LINKING_ONLY   /* ZSTD_compressionParameters */
 #include "zstd.h"     /* ZSTD_compressionParameters */
 
-int BMK_benchFiles(const char** fileNamesTable, unsigned nbFiles,const char* dictFileName, 
+int BMK_benchFiles(const char** fileNamesTable, unsigned nbFiles,const char* dictFileName,
                    int cLevel, int cLevelLast, ZSTD_compressionParameters* compressionParams);
 
 /* Set Parameters */
 void BMK_SetNbSeconds(unsigned nbLoops);
 void BMK_SetBlockSize(size_t blockSize);
-void BMK_setAdditionalParam(int additionalParam);
+void BMK_SetNbThreads(unsigned nbThreads);
 void BMK_setNotificationLevel(unsigned level);
-void BMK_setDecodeOnly(unsigned decodeFlag);
+void BMK_setAdditionalParam(int additionalParam);
+void BMK_setDecodeOnlyMode(unsigned decodeFlag);
 
 #endif   /* BENCH_H_121279284357 */
diff --git a/programs/zstdcli.c b/programs/zstdcli.c
index 978ffcfe..03ad1ac7 100644
--- a/programs/zstdcli.c
+++ b/programs/zstdcli.c
@@ -352,7 +352,7 @@ int main(int argCount, const char* argv[])
                          /* Decoding */
                     case 'd':
 #ifndef ZSTD_NOBENCH
-                            if (operation==zom_bench) { BMK_setDecodeOnly(1); argument++; break; }  /* benchmark decode (hidden option) */
+                            if (operation==zom_bench) { BMK_setDecodeOnlyMode(1); argument++; break; }  /* benchmark decode (hidden option) */
 #endif
                             operation=zom_decompress; argument++; break;
 
@@ -430,6 +430,12 @@ int main(int argCount, const char* argv[])
                         dictSelect = readU32FromChar(&argument);
                         break;
 
+                        /* nb of threads (hidden option) */
+                    case 'T':
+                        argument++;
+                        BMK_SetNbThreads(readU32FromChar(&argument));
+                        break;
+
                         /* Pause at the end (-p) or set an additional param (-p#) (hidden option) */
                     case 'p': argument++;
 #ifndef ZSTD_NOBENCH