diff --git a/src/toil_vg/vg_surject.py b/src/toil_vg/vg_surject.py
index 6553499b..019ccd29 100644
--- a/src/toil_vg/vg_surject.py
+++ b/src/toil_vg/vg_surject.py
@@ -40,6 +40,8 @@ def surject_subparser(parser):
         help="Path to xg index")    
     parser.add_argument("--paths", nargs='+', default = [],
         help="list of path names to surject to (default: all in xg)")
+    parser.add_argument("--ref_paths_file", type=make_url,
+        help="file containing reference paths for SAM header (if 2nd column present, it will be used for lengths)")
     parser.add_argument("--interleaved", action="store_true", default=False,
         help="treat gam as interleaved read pairs.  overrides map-args")
     parser.add_argument("--gam_input_reads", type=make_url, required=True,
@@ -58,7 +60,7 @@ def surject_subparser(parser):
     add_container_tool_parse_args(parser)
 
     
-def run_surjecting(job, context, gam_input_reads_id, output_name, interleaved, xg_file_id, paths):
+def run_surjecting(job, context, gam_input_reads_id, output_name, interleaved, xg_file_id, paths, ref_paths_id):
     """ split the fastq, then surject each chunk.  returns outputgams, paired with total surject time
     (excluding toil-vg overhead such as transferring and splitting files )"""
 
@@ -76,10 +78,10 @@ def run_surjecting(job, context, gam_input_reads_id, output_name, interleaved, x
         reads_chunk_ids = [[r] for r in [gam_input_reads_id]]
 
     return child_job.addFollowOnJobFn(run_whole_surject, context, reads_chunk_ids, output_name, 
-                                      interleaved, xg_file_id, paths, cores=context.config.misc_cores,
+                                      interleaved, xg_file_id, paths, ref_paths_id, cores=context.config.misc_cores,
                                       memory=context.config.misc_mem, disk=context.config.misc_disk).rv()
     
-def run_whole_surject(job, context, reads_chunk_ids, output_name, interleaved, xg_file_id, paths):
+def run_whole_surject(job, context, reads_chunk_ids, output_name, interleaved, xg_file_id, paths, ref_paths_id):
     """
     Surject all gam chunks in parallel.
     
@@ -106,7 +108,7 @@ def run_whole_surject(job, context, reads_chunk_ids, output_name, interleaved, x
     for chunk_id, chunk_filename_ids in enumerate(zip(*reads_chunk_ids)):
         #Run graph surject on each gam chunk
         chunk_surject_job = child_job.addChildJobFn(run_chunk_surject, context, interleaved, xg_file_id,
-                                                    paths, chunk_filename_ids, '{}_chunk{}'.format(output_name, chunk_id),
+                                                    paths, ref_paths_id, chunk_filename_ids, '{}_chunk{}'.format(output_name, chunk_id),
                                                     cores=context.config.alignment_cores,
                                                     memory=context.config.alignment_mem,
                                                     disk=context.config.alignment_disk)
@@ -118,7 +120,7 @@ def run_whole_surject(job, context, reads_chunk_ids, output_name, interleaved, x
                                       memory=context.config.misc_mem, disk=context.config.misc_disk).rv()
 
 
-def run_chunk_surject(job, context, interleaved, xg_file_id, paths, chunk_filename_ids, chunk_id):
+def run_chunk_surject(job, context, interleaved, xg_file_id, paths, ref_paths_id, chunk_filename_ids, chunk_id):
     """ run surject on a chunk.  interface mostly copied from run_chunk_alignment.
     
     Takes an xg file and path colleciton to surject against, a list of chunk
@@ -149,6 +151,10 @@ def run_chunk_surject(job, context, interleaved, xg_file_id, paths, chunk_filena
         gam_file = os.path.join(work_dir, 'reads_chunk_{}_{}.{}'.format(chunk_id, j, reads_ext))
         job.fileStore.readGlobalFile(chunk_filename_id, gam_file)
         gam_files.append(gam_file)
+
+    if ref_paths_id:
+        ref_paths_file = os.path.join(work_dir, 'ref-paths.tsv')
+        job.fileStore.readGlobalFile(ref_paths_id, ref_paths_file)
     
     # And a temp file for our surject output
     output_file = os.path.join(work_dir, "surject_{}.bam".format(chunk_id))
@@ -162,8 +168,10 @@ def run_chunk_surject(job, context, interleaved, xg_file_id, paths, chunk_filena
         cmd += ['-x', os.path.basename(xg_file)]
         for surject_path in paths:
             cmd += ['--into-path', surject_path]
+        if ref_paths_id:
+            cmd += ['--ref-paths', os.path.basename(ref_paths_file)]
         cmd += ['-t', str(context.config.alignment_cores)]
-        
+
         # Mark when we start the surjection
         start_time = timeit.default_timer()
         try:
@@ -244,6 +252,7 @@ def surject_main(context, options):
             # Upload local files to the remote IO Store
             inputXGFileID = importer.load(options.xg_index)
             inputGAMFileID = importer.load(options.gam_input_reads)
+            inputRefPathsID = importer.load(options.ref_paths_file) if options.ref_paths_file else None
 
             importer.wait()
 
@@ -251,6 +260,7 @@ def surject_main(context, options):
             root_job = Job.wrapJobFn(run_surjecting, context, importer.resolve(inputGAMFileID), 'surject',
                                      options.interleaved,
                                      importer.resolve(inputXGFileID), options.paths,
+                                     importer.resolve(inputRefPathsID) if options.ref_paths_file else None,
                                      cores=context.config.misc_cores,
                                      memory=context.config.misc_mem,
                                      disk=context.config.misc_disk)