diff --git a/molecularnodes/io/download.py b/molecularnodes/io/download.py
new file mode 100644
index 00000000..069fdc1c
--- /dev/null
+++ b/molecularnodes/io/download.py
@@ -0,0 +1,80 @@
+import os
+import requests
+import io
+
+def fetch(code, format="cif", cache=None, database='rcsb'):
+    """
+    Downloads a structure from the specified protein data bank in the given format.
+
+    Parameters
+    ----------
+    code : str
+        The code of the file to fetch.
+    format : str, optional
+        The format of the file. Defaults to "cif". Possible values are ['cif', 'pdb', 
+        'mmcif', 'pdbx', 'mmtf', 'bcif'].
+    cache : str, optional
+        The cache directory to store the fetched file. Defaults to None.
+    database : str, optional
+        The database to fetch the file from. Defaults to 'rcsb'.
+
+    Returns
+    -------
+    file
+        The fetched file as a file-like object.
+
+    Raises
+    ------
+    ValueError
+        If the specified format is not supported.
+    """
+    supported_formats = ['cif', 'pdb', 'mmtf', 'bcif']
+    if format not in supported_formats:
+        raise ValueError(f"File format '{format}' not in: {supported_formats=}")
+
+    _is_binary = (format in ['bcif', 'mmtf'])
+    filename = f"{code}.{format}"
+    # create the cache location
+    if cache:
+        if not os.path.isdir(cache):
+            os.makedirs(cache)
+
+        file = os.path.join(cache, filename)
+    else:
+        file = None
+
+    # get the contents of the url
+    r = requests.get(_url(code, format, database))
+    if _is_binary:
+        content = r.content
+    else:
+        content = r.text
+
+    if file:
+        mode = "wb+" if _is_binary else "w+"
+        with open(file, mode) as f:
+            f.write(content)
+    else:
+        if _is_binary:
+            file = io.BytesIO(content)
+        else:
+            file = io.StringIO(content)
+
+    return file
+
+def _url(code, format, database="rcsb"):
+    "Get the URL for downloading the given file form a particular database."
+    
+    if database == "rcsb":
+        if format == "bcif":
+            return f"https://models.rcsb.org/{code}.bcif"
+        if format == "mmtf":
+            return f"https://mmtf.rcsb.org/v1.0/full/{code}"
+            
+        else:
+            return f"https://files.rcsb.org/download/{code}.{format}"
+    # if database == "pdbe":
+    #     return f"https://www.ebi.ac.uk/pdbe/entry-files/download/{filename}"
+    else:
+        ValueError(f"Database {database} not currently supported.")
+
diff --git a/tests/test_download.py b/tests/test_download.py
new file mode 100644
index 00000000..c43f16b8
--- /dev/null
+++ b/tests/test_download.py
@@ -0,0 +1,81 @@
+import os
+import io
+import pytest
+from molecularnodes.io import download
+import biotite.database.rcsb as rcsb
+from biotite.structure.io import load_structure
+import tempfile
+
+from .constants import codes
+
+databases = ['rcsb'] # currently can't figure out downloading from other services
+
+def _filestart(format):
+    if format == "cif":
+        return 'data_'
+    else:
+        return 'HEADER'
+
+
+
+@pytest.mark.parametrize('format', ['cif', 'mmtf', 'pdb'])
+def test_compare_biotite(format):
+    struc_download = load_structure(download.fetch('4ozs', format=format, cache=tempfile.TemporaryDirectory().name))
+    struc_biotite  = load_structure(rcsb.fetch('4ozs', format=format, target_path=tempfile.TemporaryDirectory().name))
+    assert struc_download == struc_biotite
+
+@pytest.mark.parametrize('code', codes)
+@pytest.mark.parametrize('database', databases)
+@pytest.mark.parametrize('format', ['pdb', 'cif'])
+def test_fetch_with_cache(tmpdir, code, format, database):
+    cache_dir = tmpdir.mkdir("cache")
+    file = download.fetch(code, format, cache=str(cache_dir), database=database)
+    
+    assert isinstance(file, str)
+    assert os.path.isfile(file)
+    assert file.endswith(f"{code}.{format}")
+    
+    
+    with open(file, "r") as f:
+        content = f.read()
+    assert content.startswith(_filestart(format))
+
+databases = ['rcsb'] # currently can't figure out downloading from the pdbe
+
+@pytest.mark.parametrize('code', codes)
+@pytest.mark.parametrize('database', databases)
+@pytest.mark.parametrize('format', ['pdb', 'cif'])
+def test_fetch_without_cache(tmpdir, code, format, database):
+    file = download.fetch(code, format, cache=None, database=database)
+    
+    assert isinstance(file, io.StringIO)
+    content = file.getvalue()
+    assert content.startswith(_filestart(format))
+
+@pytest.mark.parametrize('database', databases)
+def test_fetch_with_invalid_format(database):
+    code = '4OZS'
+    format = "xyz"
+    
+    with pytest.raises(ValueError):
+        download.fetch(code, format, cache=None, database=database)
+
+@pytest.mark.parametrize('code', codes)
+@pytest.mark.parametrize('database', databases)
+@pytest.mark.parametrize('format', ['bcif', 'mmtf'])
+def test_fetch_with_binary_format(tmpdir, code, database, format):
+    cache_dir = tmpdir.mkdir("cache")
+    file = download.fetch(code, format, cache=str(cache_dir), database=database)
+    
+    assert isinstance(file, str)
+    assert os.path.isfile(file)
+    assert file.endswith(f"{code}.{format}")
+    
+    if format == "bcif":
+        start = b"\x83\xa7"
+    elif format == "mmtf":
+        start = b"\xde\x00"
+    
+    with open(file, "rb") as f:
+        content = f.read()
+    assert content.startswith(start)
\ No newline at end of file