diff --git a/reader.py b/reader.py
index 5845775..cbe0fc5 100644
--- a/reader.py
+++ b/reader.py
@@ -1,15 +1,23 @@
 import os
+import sys
 import time
 import logging
 import requests
+import threading
+from copy import copy
+from Queue import Queue
+from datetime import datetime
 from homura import download as fetch
 from tempfile import mkdtemp
-from datetime import datetime
 from collections import OrderedDict
 
 logger = logging.getLogger('landsat8.meta')
 
 
+class ReachedEndOfProcess(Exception):
+    pass
+
+
 def convert_date(value):
     return datetime.strptime(value, '%Y-%m-%d').date()
 
@@ -28,11 +36,27 @@ def download_meta(url, download_path):
     return open(dpath, 'r')
 
 
+def row_processor(record, date, dst, writers):
+
+    path = os.path.join(dst, str(date.year), str(date.month), str(date.day))
+
+    logger.info('processing %s' % record['sceneID'])
+    for w in writers:
+        w(path, record)
+
+
 def csv_reader(dst, writers, start_date=None, end_date=None, url=None,
-               download=False, download_path=None):
+               download=False, download_path=None, num_worker_threads=1):
     """ Reads landsat8 metadata from a csv file stored on USGS servers
     and applys writer functions on the data """
 
+    threaded = False
+    threads = []
+
+    if num_worker_threads > 0:
+        threaded = True
+        queue = Queue()
+
     if not url:
         url = 'http://landsat.usgs.gov/metadata_service/bulk_metadata_files/LANDSAT_8.csv'
 
@@ -50,43 +74,72 @@ def csv_reader(dst, writers, start_date=None, end_date=None, url=None,
         r = requests.get(url, stream=True)
         liner = r.iter_lines
 
+    if start_date:
+        start_date = convert_date(start_date)
+
+    if end_date:
+        end_date = convert_date(end_date)
+
     header = None
-    start_write = False
 
+    counter = 0
     for line in liner():
         row = line.split(',')
 
         # first line is the header
         if not header:
             header = row
+            continue
 
         # other lines have values
         else:
-            for j, v in enumerate(row):
-                try:
-                    row[j] = float(v)
-                except ValueError:
-                    pass
-
-            # generate the record
-            record = OrderedDict(zip(header, row))
-
-            # apply filter
-            # if there is an enddate, stops the process when the end date is reached
-            if not end_date:
-                start_write = True
-
-            if end_date and record['acquisitionDate'] == end_date:
-                start_write = True
-
-            if start_date and record['acquisitionDate'] == start_date:
-                break
-
-            # if condition didn't match, generate path and apply writers and go to the next line
-            if start_write:
-                date = convert_date(record['acquisitionDate'])
-                path = os.path.join(dst, str(date.year), str(date.month), str(date.day))
-
-                logger.info('processing %s' % record['sceneID'])
-                for w in writers:
-                    w(path, record)
+            write = False
+
+        for j, v in enumerate(row):
+            try:
+                row[j] = float(v)
+            except ValueError:
+                pass
+
+        # generate the record
+        record = OrderedDict(zip(header, row))
+        date = convert_date(record['acquisitionDate'])
+
+        # apply filter
+        # if there is an enddate, stops the process when the end date is reached
+        if not end_date:
+            write = True
+
+        if end_date and date <= end_date:
+            write = True
+
+        if start_date and date < start_date:
+            break
+
+        if write:
+            if threaded:
+                new_record = copy(record)
+                queue.put([new_record, date, dst, writers])
+                counter += 1
+            else:
+                row_processor(record, date, dst, writers)
+
+            if threaded and counter > 500:
+
+                def worker():
+                    while not queue.empty():
+                        args = queue.get()
+                        try:
+                            row_processor(*args)
+                        except Exception:
+                            exc = sys.exc_info()
+                            logger.error('%s | %s scene skipped due to error: %s' % (threading.current_thread().name,
+                                                                                     args[0]['sceneID'],
+                                                                                     exc[1].__str__()))
+                        queue.task_done()
+
+                if not threads:
+                    for i in range(num_worker_threads):
+                        t = threading.Thread(target=worker)
+                        t.start()
+                        threads.append(t)