diff --git a/src/openfe_analysis/cli.py b/src/openfe_analysis/cli.py
index 42826cc..7e761b6 100644
--- a/src/openfe_analysis/cli.py
+++ b/src/openfe_analysis/cli.py
@@ -1,8 +1,12 @@
 import click
+from click.exceptions import BadOptionUsage
 import json
+import netCDF4 as nc
 import pathlib
+import tqdm
 
 from . import rmsd
+from .reader import FEReader
 
 
 @click.group()
@@ -20,6 +24,7 @@ def cli():
                                           dir_okay=False,
                                           path_type=pathlib.Path))
 def rfe_analysis(loc, output):
+    """Perform structural analysis on OpenMM RFE simulation"""
     pdb = loc / "hybrid_system.pdb"
     trj = loc / "simulation.nc"
 
@@ -27,3 +32,75 @@ def rfe_analysis(loc, output):
 
     with click.open_file(output, 'w') as f:
         f.write(json.dumps(data))
+
+
+_statehelp = """\
+"""
+
+
+@cli.command(name='RFE_trjconv')
+@click.argument("loc", type=click.Path(exists=True,
+                                       readable=True,
+                                       file_okay=False,
+                                       dir_okay=True,
+                                       path_type=pathlib.Path))
+@click.argument('output', type=click.Path(writable=True,
+                                        dir_okay=False,
+                                        exists=False,
+                                        path_type=pathlib.Path),
+                        required=True)
+@click.option('-s', '--state', required=True,
+              help="either an integer (0 and -1 giving endstates) or 'all'")
+def RFE_trjconv(loc, output, state):
+    """Convert .nc trajectory files from RBFE to new format for a single state
+
+    LOC is the directory where a simulation took place, it should contain the
+    simulation.nc and hybrid_system.pdb files that were produced.
+
+    OUTPUT is the name of the new trajectory file, e.g. "out.xtc".  Any file
+    format supported by MDAnalysis can be specified, including XTC and DCD
+    formats.
+
+    The .nc trajectory file contains multiple states; a single state can be
+    specified for output.  Negative indices are allowed and treated as in
+    Python, therefore ``--state=0`` or ``--state=-1`` will produce trajectories
+    of the two end states.
+
+    If ``--state='all'`` is given, all states are outputted, and the output
+    filename has the state number inserted before the file prefix,
+    e.g. ``--output=traj.dcd`` would produce a files called ``traj_state0.dcd``
+    etc.
+    """
+    import MDAnalysis as mda
+
+    pdb = loc / "hybrid_system.pdb"
+    trj = loc / "simulation.nc"
+
+    ds = nc.Dataset(trj, mode='r')
+
+    if state == 'all':
+        # figure out how many states we need to output
+        nstates = ds.dimensions['state'].size
+
+        states = range(nstates)
+        # turn out.dcd -> out_0.dcd
+        outputs = [
+            output.with_stem(output.stem + f'_state{i}')
+            for i in range(nstates)
+        ]
+    else:
+        try:
+            states = [int(state)]
+        except ValueError:
+            raise BadOptionUsage(f"Invalid state specified: {state}")
+        outputs = [output]
+
+    for s, o in zip(states, outputs):
+        u = mda.Universe(pdb, ds,
+                         format=FEReader,
+                         state_id=s)
+        ag = u.atoms  # todo, atom selections would be here
+
+        with mda.Writer(str(o), n_atoms=len(ag)) as w:
+            for ts in tqdm.tqdm(u.trajectory):
+                w.write(ag)
diff --git a/src/openfe_analysis/reader.py b/src/openfe_analysis/reader.py
index 5eebf14..dd64242 100644
--- a/src/openfe_analysis/reader.py
+++ b/src/openfe_analysis/reader.py
@@ -55,12 +55,12 @@ def __init__(self, filename, convert_units=True, **kwargs):
         convert_units : bool
           convert positions to A
         """
-        self._state_id = kwargs.pop('state_id', None)
-        self._replica_id = kwargs.pop('replica_id', None)
-        if not ((self._state_id is None) ^ (self._replica_id is None)):
+        s_id = kwargs.pop('state_id', None)
+        r_id = kwargs.pop('replica_id', None)
+        if not ((s_id is None) ^ (r_id is None)):
             raise ValueError("Specify one and only one of state or replica, "
-                             f"got state id={self._state_id} "
-                             f"replica_id={self._replica_id}")
+                             f"got state id={s_id} "
+                             f"replica_id={r_id}")
 
         super().__init__(filename, convert_units, **kwargs)
 
@@ -70,6 +70,15 @@ def __init__(self, filename, convert_units=True, **kwargs):
         else:
             self._dataset = nc.Dataset(filename)
             self._dataset_owner = True
+
+        # if we have a negative indexed state_id or replica_id, convert this
+        if s_id is not None and s_id < 0:
+            s_id = range(self._dataset.dimensions['state'].size)[s_id]
+        elif r_id is not None and r_id < 0:
+            r_id = range(self._dataset.dimensions['replica'].size)[r_id]
+        self._state_id = s_id
+        self._replica_id = r_id
+
         self._n_atoms = self._dataset.dimensions['atom'].size
         self.ts = Timestep(self._n_atoms)
         self._dt = _determine_dt(self._dataset)
diff --git a/src/tests/test_reader.py b/src/tests/test_reader.py
index fe3f997..7b36225 100644
--- a/src/tests/test_reader.py
+++ b/src/tests/test_reader.py
@@ -24,3 +24,17 @@ def test_universe_from_nc_file(simulation_nc, hybrid_system_pdb):
     assert len(u.atoms) == 4782
 
 
+def test_universe_creation_negative_state(simulation_nc, hybrid_system_pdb):
+    u = mda.Universe(hybrid_system_pdb, simulation_nc,
+                     format='openfe rfe', state_id=-1)
+
+    assert u.trajectory._state_id == 10
+    assert u.trajectory._replica_id is None
+
+
+def test_universe_creation_negative_replica(simulation_nc, hybrid_system_pdb):
+    u = mda.Universe(hybrid_system_pdb, simulation_nc,
+                     format='openfe rfe', replica_id=-1)
+
+    assert u.trajectory._state_id is None
+    assert u.trajectory._replica_id == 10