diff --git a/Makefile b/Makefile
index 4719af9..1d234a8 100644
--- a/Makefile
+++ b/Makefile
@@ -1,4 +1,6 @@
 year=$(shell date +%Y)
+START_TIME=$(shell export TZ=UTC; date -Iseconds)
+TIME_LIMIT=21600
 
 .PHONY: all
 all: upload
@@ -54,11 +56,12 @@ CIVIL_SCRAPE_START_QUERY=$(shell tail -n +2 scripts/nightly_civil_start.sql)
 civil-%.jl: cases.db
 	START=$$(sqlite-utils query --csv --no-headers cases.db \
 	      "$(CIVIL_SCRAPE_START_QUERY)" -p subdivision $*); \
-				echo $$START; \
+	      export START_TIME=$(START_TIME); export TIME_LIMIT=$(TIME_LIMIT); \
 	      scrapy crawl civil -s CLOSESPIDER_TIMEOUT=3600 -a year=$(year) -a division=$* -a start=$$START -O $@;
 
 chancery.jl: cases.db
 	START=$$(sqlite3 cases.db < scripts/nightly_chancery_start.sql); \
+	      export START_TIME=$(START_TIME); export TIME_LIMIT=$(TIME_LIMIT); \
 	      scrapy crawl chancery -a year=$(year) -a start=$$START -O $@;
 
 cases.db :
diff --git a/courtscraper/spiders/base.py b/courtscraper/spiders/base.py
index 73aa10a..0babffb 100644
--- a/courtscraper/spiders/base.py
+++ b/courtscraper/spiders/base.py
@@ -1,4 +1,6 @@
+import os
 from abc import ABC, abstractmethod
+from datetime import datetime, timedelta, timezone
 
 from scrapy import Spider
 from scrapy.exceptions import CloseSpider
@@ -27,6 +29,14 @@ def __init__(
         else:
             self.case_numbers = self.get_case_numbers(self.year)
 
+        start_time_iso = os.getenv(
+            "START_TIME", datetime.now(tz=timezone.utc).isoformat()
+        )
+        self.start_time = datetime.fromisoformat(start_time_iso)
+
+        time_limit_in_secs = os.getenv("TIME_LIMIT", 21600)
+        self.time_limit = timedelta(seconds=int(time_limit_in_secs))
+
         super().__init__(**kwargs)
 
     @property
@@ -47,6 +57,19 @@ def start_requests(self):
     def get_case_numbers(self):
         pass
 
+    def out_of_time(self) -> bool:
+        """
+        Checks whether the we have enough time to continue scraping.
+        We'll assume we need at most 30 minutes to clean up and finish
+        post-scrape tasks.
+        """
+
+        runtime = datetime.now(tz=timezone.utc) - self.start_time
+        if runtime >= self.time_limit:
+            return True
+
+        return False
+
     def case_numbers_from_file(self, filename):
         with open(filename) as f:
             for case_number in f:
diff --git a/courtscraper/spiders/chancery.py b/courtscraper/spiders/chancery.py
index a014f9e..b1dcec2 100644
--- a/courtscraper/spiders/chancery.py
+++ b/courtscraper/spiders/chancery.py
@@ -21,6 +21,9 @@ def get_case_numbers(self, year):
 
     def start_requests(self):
         for case_number in self.case_numbers:
+            if self.out_of_time():
+                break
+
             yield Request(
                 ChancerySpider.url,
                 meta={
diff --git a/courtscraper/spiders/civil.py b/courtscraper/spiders/civil.py
index d1db6d7..750750c 100644
--- a/courtscraper/spiders/civil.py
+++ b/courtscraper/spiders/civil.py
@@ -13,6 +13,9 @@ def __init__(self, division="2", **kwargs):
 
     def start_requests(self):
         for case_number in self.case_numbers:
+            if self.out_of_time():
+                break
+
             yield Request(
                 CivilSpider.url,
                 meta={