diff --git a/gstlal-inspiral/bin/gstlal_inspiral b/gstlal-inspiral/bin/gstlal_inspiral
index dd50179573480a45da0a95d023d30b79e612a47c..a5d75883b16449115595adfc9e26da0ee6e2b658 100755
--- a/gstlal-inspiral/bin/gstlal_inspiral
+++ b/gstlal-inspiral/bin/gstlal_inspiral
@@ -178,6 +178,7 @@ import socket
 import time
 import tempfile
 import math
+import subprocess
 from collections import namedtuple
 
 # The following snippet is taken from http://gstreamer.freedesktop.org/wiki/FAQ#Mypygstprogramismysteriouslycoredumping.2Chowtofixthis.3F
@@ -546,7 +547,22 @@ else:
 #
 
 
-for svd_bank, output_filename, likelihood_namedtuple in zip(svd_banks, options.output, likelihood_namedtuples_list):
+for output_file_number, (svd_bank, output_filename, likelihood_namedtuple) in enumerate(zip(svd_banks, options.output, likelihood_namedtuples_list)):
+	#
+	# Checkpointing only supported for gzip files in offline analysis
+	# FIXME Is there a better way to check if a file is corrupted? Implement a means by which to check for sqlite file validity
+	#
+
+	if options.data_source not in ("lvshm", "framexmit") and output_filename.endswith('.gz'):
+		try:
+			subprocess.check_call(["gzip", "--test", output_filename])
+			# File is OK and there is no need to process it, skip ahead in the loop
+			continue
+		except subprocess.CalledProcessError:
+			# File does not exist or is corrupted, need to reprocess
+			print >>sys.stderr, "Checkpoint: {0} of {1} files completed and continuing with {2}".format(output_file_number, len(options.output), os.path.basename(output_filename))
+			pass
+
 	#
 	# create a new, empty, Bottle application and make it the current
 	# default, then start http server(s) to serve it up