diff --git a/tests/Makefile.am b/tests/Makefile.am
index 86084e17..4116b81a 100644
--- a/tests/Makefile.am
+++ b/tests/Makefile.am
@@ -7,6 +7,7 @@
 
 EXTRA_DIST = \
 	files \
+	ossfuzz \
 	tests.h \
 	test_files.sh \
 	test_compress.sh \
diff --git a/tests/ossfuzz/Makefile b/tests/ossfuzz/Makefile
new file mode 100644
index 00000000..67ddbe5a
--- /dev/null
+++ b/tests/ossfuzz/Makefile
@@ -0,0 +1,7 @@
+fuzz: fuzz.c
+	$(CC) $(CFLAGS) -c fuzz.c -I ../../src/liblzma/api/
+	$(CXX) $(CXXFLAGS) -lFuzzingEngine fuzz.o -o $(OUT)/fuzz \
+	      ../../src/liblzma/.libs/liblzma.a
+
+clean:
+	rm -f *.o
diff --git a/tests/ossfuzz/config/fuzz.dict b/tests/ossfuzz/config/fuzz.dict
new file mode 100644
index 00000000..932d67c2
--- /dev/null
+++ b/tests/ossfuzz/config/fuzz.dict
@@ -0,0 +1,2 @@
+"\xFD7zXZ\x00"
+"YZ"
diff --git a/tests/ossfuzz/config/fuzz.options b/tests/ossfuzz/config/fuzz.options
new file mode 100644
index 00000000..d59dfc13
--- /dev/null
+++ b/tests/ossfuzz/config/fuzz.options
@@ -0,0 +1,2 @@
+[libfuzzer]
+dict = fuzz.dict
diff --git a/tests/ossfuzz/fuzz.c b/tests/ossfuzz/fuzz.c
new file mode 100644
index 00000000..6d899302
--- /dev/null
+++ b/tests/ossfuzz/fuzz.c
@@ -0,0 +1,82 @@
+///////////////////////////////////////////////////////////////////////////////
+//
+/// \file       fuzz.c
+/// \brief      Fuzz test program for liblzma
+//
+//  Author:     Lasse Collin
+//
+//  This file has been put into the public domain.
+//  You can do whatever you want with this file.
+//
+///////////////////////////////////////////////////////////////////////////////
+
+#include <inttypes.h>
+#include <stdlib.h>
+#include <stdio.h>
+#include "lzma.h"
+
+
+// Output buffer for decompressed data. This is write only; nothing cares
+// about the actual data written here.
+static uint8_t outbuf[4096];
+
+
+extern int
+LLVMFuzzerTestOneInput(const uint8_t *inbuf, size_t inbuf_size)
+{
+	// Some header values can make liblzma allocate a lot of RAM
+	// (up to about 4 GiB with liblzma 5.2.x). We set a limit here to
+	// prevent extreme allocations when fuzzing.
+	const uint64_t memlimit = 300 << 20; // 300 MiB
+
+	// Initialize a .xz decoder using the above memory usage limit.
+	// Enable support for concatenated .xz files which is used when
+	// decompressing regular .xz files (instead of data embedded inside
+	// some other file format). Integrity checks on the uncompressed
+	// data are ignored to make fuzzing more effective (incorrect check
+	// values won't prevent the decoder from processing more input).
+	//
+	// The flag LZMA_IGNORE_CHECK doesn't disable verification of header
+	// CRC32 values. Those checks are disabled when liblzma is built
+	// with the #define FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION.
+	lzma_stream strm = LZMA_STREAM_INIT;
+	lzma_ret ret = lzma_stream_decoder(&strm, memlimit,
+			LZMA_CONCATENATED | LZMA_IGNORE_CHECK);
+	if (ret != LZMA_OK) {
+		// This should never happen unless the system has
+		// no free memory or address space to allow the small
+		// allocations that the initialization requires.
+		fprintf(stderr, "lzma_stream_decoder() failed (%d)\n", ret);
+		abort();
+	}
+
+	// Give the whole input buffer at once to liblzma.
+	// Output buffer isn't initialized as liblzma only writes to it.
+	strm.next_in = inbuf;
+	strm.avail_in = inbuf_size;
+	strm.next_out = outbuf;
+	strm.avail_out = sizeof(outbuf);
+
+	while ((ret = lzma_code(&strm, LZMA_FINISH)) == LZMA_OK) {
+		if (strm.avail_out == 0) {
+			// outbuf became full. We don't care about the
+			// uncompressed data there, so we simply reuse
+			// the outbuf and overwrite the old data.
+			strm.next_out = outbuf;
+			strm.avail_out = sizeof(outbuf);
+		}
+	}
+
+	// LZMA_PROG_ERROR should never happen as long as the code calling
+	// the liblzma functions is correct. Thus LZMA_PROG_ERROR is a sign
+	// of a bug in either this function or in liblzma.
+	if (ret == LZMA_PROG_ERROR) {
+		fprintf(stderr, "lzma_code() returned LZMA_PROG_ERROR\n");
+		abort();
+	}
+
+	// Free the allocated memory.
+	lzma_end(&strm);
+
+	return 0;
+}