diff --git a/datasette/app.py b/datasette/app.py
index 7502009a45..72ff526435 100644
--- a/datasette/app.py
+++ b/datasette/app.py
@@ -18,7 +18,6 @@
 import time
 from .utils import (
     Filters,
-    compound_pks_from_path,
     CustomJSONEncoder,
     compound_keys_after_sql,
     detect_fts_sql,
@@ -33,6 +32,7 @@
     path_with_ext,
     sqlite_timelimit,
     to_css_class,
+    urlsafe_components,
     validate_sql_select,
 )
 from .version import __version__
@@ -613,6 +613,14 @@ async def data(self, request, name, hash, table):
             search_description = 'search matches "{}"'.format(search)
             params['search'] = search
 
+        # Allow for custom sort order
+        sort = special_args.get('_sort')
+        if sort:
+            order_by = sort
+        sort_desc = special_args.get('_sort_desc')
+        if sort_desc:
+            order_by = '{} desc'.format(sort_desc)
+
         count_sql = 'select count(*) from {table_name} {where}'.format(
             table_name=escape_sqlite(table),
             where=(
@@ -638,20 +646,46 @@ async def data(self, request, name, hash, table):
             if is_view:
                 # _next is an offset
                 offset = ' offset {}'.format(int(_next))
-            elif use_rowid:
-                where_clauses.append(
-                    'rowid > :p{}'.format(
-                        len(params),
-                    )
-                )
-                params['p{}'.format(len(params))] = _next
             else:
-                pk_values = compound_pks_from_path(_next)
-                if len(pk_values) == len(pks):
-                    param_len = len(params)
-                    where_clauses.append(compound_keys_after_sql(pks, param_len))
-                    for i, pk_value in enumerate(pk_values):
-                        params['p{}'.format(param_len + i)] = pk_value
+                components = urlsafe_components(_next)
+                # If a sort order is applied, the first of these is the sort value
+                if sort or sort_desc:
+                    sort_value = components[0]
+                    components = components[1:]
+                    print('sort_varlue = {}, components = {}'.format(
+                        sort_value, components
+                    ))
+
+                # Figure out the SQL for next-based-on-primary-key first
+                next_by_pk_clauses = []
+                if use_rowid:
+                    next_by_pk_clauses.append(
+                        'rowid > :p{}'.format(
+                            len(params),
+                        )
+                    )
+                    params['p{}'.format(len(params))] = components[0]
+                else:
+                    # Apply the tie-breaker based on primary keys
+                    if len(components) == len(pks):
+                        param_len = len(params)
+                        next_by_pk_clauses.append(compound_keys_after_sql(pks, param_len))
+                        for i, pk_value in enumerate(components):
+                            params['p{}'.format(param_len + i)] = pk_value
+
+                # Now add the sort SQL, which may incorporate next_by_pk_clauses
+                if sort or sort_desc:
+                    where_clauses.append(
+                        '({column} {op} :p{p} or ({column} = :p{p} and {next_clauses}))'.format(
+                            column=escape_sqlite(sort or sort_desc),
+                            op='>' if sort else '<',
+                            p=len(params),
+                            next_clauses=' and '.join(next_by_pk_clauses),
+                        )
+                    )
+                    params['p{}'.format(len(params))] = sort_value
+                else:
+                    where_clauses.extend(next_by_pk_clauses)
 
         where_clause = ''
         if where_clauses:
@@ -707,9 +741,26 @@ async def data(self, request, name, hash, table):
                 next_value = int(_next or 0) + self.page_size
             else:
                 next_value = path_from_row_pks(rows[-2], pks, use_rowid)
-            next_url = urllib.parse.urljoin(request.url, path_with_added_args(request, {
-                '_next': next_value,
-            }))
+            # If there's a sort or sort_desc, add that value as a prefix
+            if (sort or sort_desc) and not is_view:
+                prefix = str(rows[-2][sort or sort_desc])
+                next_value = '{},{}'.format(
+                    urllib.parse.quote_plus(prefix), next_value
+                )
+                added_args = {
+                    '_next': next_value,
+                }
+                if sort:
+                    added_args['_sort'] = sort
+                else:
+                    added_args['_sort_desc'] = sort_desc
+            else:
+                added_args = {
+                    '_next': next_value,
+                }
+            next_url = urllib.parse.urljoin(request.url, path_with_added_args(
+                request, added_args
+            ))
             rows = rows[:self.page_size]
 
         # Number of filtered rows in whole set:
@@ -778,7 +829,7 @@ async def extra_template():
 class RowView(RowTableShared):
     async def data(self, request, name, hash, table, pk_path):
         table = urllib.parse.unquote_plus(table)
-        pk_values = compound_pks_from_path(pk_path)
+        pk_values = urlsafe_components(pk_path)
         pks = await self.pks_for_table(name, table)
         use_rowid = not pks
         select = '*'
diff --git a/datasette/utils.py b/datasette/utils.py
index c674349b82..b7798bf45e 100644
--- a/datasette/utils.py
+++ b/datasette/utils.py
@@ -29,9 +29,10 @@
 ).split())
 
 
-def compound_pks_from_path(path):
+def urlsafe_components(token):
+    "Splits token on commas and URL decodes each component"
     return [
-        urllib.parse.unquote_plus(b) for b in path.split(',')
+        urllib.parse.unquote_plus(b) for b in token.split(',')
     ]
 
 
diff --git a/tests/fixtures.py b/tests/fixtures.py
index cf19fa3148..4ede6a906b 100644
--- a/tests/fixtures.py
+++ b/tests/fixtures.py
@@ -1,6 +1,7 @@
 from datasette.app import Datasette
 import itertools
 import os
+import random
 import sqlite3
 import sys
 import string
@@ -34,6 +35,25 @@ def generate_compound_rows(num):
         yield a, b, c, '{}-{}-{}'.format(a, b, c)
 
 
+def generate_sortable_rows(num):
+    rand = random.Random(42)
+    for a, b in itertools.islice(
+        itertools.product(string.ascii_lowercase, repeat=2), num
+    ):
+        yield {
+            'pk1': a,
+            'pk2': b,
+            'content': '{}-{}'.format(a, b),
+            'sortable': rand.randint(-100, 100),
+            'sortable_with_nulls': rand.choice([
+                None, rand.random(), rand.random()
+            ]),
+            'sortable_with_nulls_2': rand.choice([
+                None, rand.random(), rand.random()
+            ]),
+        }
+
+
 METADATA = {
     'title': 'Datasette Title',
     'description': 'Datasette Description',
@@ -70,7 +90,6 @@ def generate_compound_rows(num):
 
 INSERT INTO compound_primary_key VALUES ('a', 'b', 'c');
 
-
 CREATE TABLE compound_three_primary_keys (
   pk1 varchar(30),
   pk2 varchar(30),
@@ -79,6 +98,15 @@ def generate_compound_rows(num):
   PRIMARY KEY (pk1, pk2, pk3)
 );
 
+CREATE TABLE sortable (
+  pk1 varchar(30),
+  pk2 varchar(30),
+  content text,
+  sortable integer,
+  sortable_with_nulls real,
+  sortable_with_nulls_2 real,
+  PRIMARY KEY (pk1, pk2)
+);
 
 CREATE TABLE no_primary_key (
   content text,
@@ -142,6 +170,13 @@ def generate_compound_rows(num):
     'INSERT INTO compound_three_primary_keys VALUES ("{a}", "{b}", "{c}", "{content}");'.format(
         a=a, b=b, c=c, content=content
     ) for a, b, c, content in generate_compound_rows(1001)
+]) + '\n'.join([
+    '''INSERT INTO sortable VALUES (
+        "{pk1}", "{pk2}", "{content}", {sortable},
+        {sortable_with_nulls}, {sortable_with_nulls_2});
+    '''.format(
+        **row
+    ).replace('None', 'null') for row in generate_sortable_rows(201)
 ])
 
 
diff --git a/tests/test_api.py b/tests/test_api.py
index 733727c673..acdd5e7132 100644
--- a/tests/test_api.py
+++ b/tests/test_api.py
@@ -1,6 +1,7 @@
 from .fixtures import (
     app_client,
     generate_compound_rows,
+    generate_sortable_rows,
 )
 import pytest
 
@@ -13,7 +14,7 @@ def test_homepage(app_client):
     assert response.json.keys() == {'test_tables': 0}.keys()
     d = response.json['test_tables']
     assert d['name'] == 'test_tables'
-    assert d['tables_count'] == 9
+    assert d['tables_count'] == 10
 
 
 def test_database_page(app_client):
@@ -106,6 +107,16 @@ def test_database_page(app_client):
             'outgoing': [],
         },
         'label_column': None,
+    }, {
+        'columns': [
+            'pk1', 'pk2', 'content', 'sortable', 'sortable_with_nulls',
+            'sortable_with_nulls_2'
+        ],
+        'name': 'sortable',
+        'count': 201,
+        'hidden': False,
+        'foreign_keys': {'incoming': [], 'outgoing': []},
+        'label_column': None,
     }, {
         'columns': ['pk', 'content'],
         'name': 'table/with/slashes.csv',
@@ -345,6 +356,30 @@ def test_paginate_compound_keys_with_extra_filters(app_client):
     assert expected == [f['content'] for f in fetched]
 
 
+@pytest.mark.parametrize('query_string,sort_key', [
+    ('_sort=sortable', lambda row: row['sortable']),
+    ('_sort_desc=sortable', lambda row: -row['sortable']),
+])
+def test_sortable(app_client, query_string, sort_key):
+    path = '/test_tables/sortable.jsono?{}'.format(query_string)
+    fetched = []
+    page = 0
+    while path:
+        page += 1
+        assert page < 100
+        response = app_client.get(path, gather_request=False)
+        fetched.extend(response.json['rows'])
+        path = response.json['next_url']
+    assert 5 == page
+    expected = list(generate_sortable_rows(201))
+    expected.sort(key=sort_key)
+    assert [
+        r['content'] for r in expected
+    ] == [
+        r['content'] for r in fetched
+    ]
+
+
 @pytest.mark.parametrize('path,expected_rows', [
     ('/test_tables/simple_primary_key.json?content=hello', [
         ['1', 'hello'],
diff --git a/tests/test_utils.py b/tests/test_utils.py
index a52042561c..d71ecc60fd 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -18,8 +18,8 @@
     ('123%2C433,112', ['123,433', '112']),
     ('123%2F433%2F112', ['123/433/112']),
 ])
-def test_compound_pks_from_path(path, expected):
-    assert expected == utils.compound_pks_from_path(path)
+def test_urlsafe_components(path, expected):
+    assert expected == utils.urlsafe_components(path)
 
 
 @pytest.mark.parametrize('row,pks,expected_path', [