diff --git a/yt/frontends/ramses/io.py b/yt/frontends/ramses/io.py
index 6f241631a8..7cbfab8667 100644
--- a/yt/frontends/ramses/io.py
+++ b/yt/frontends/ramses/io.py
@@ -144,7 +144,8 @@ def _ramses_particle_csv_file_handler(particle_handler, subset, fields, count):
     # read only selected fields
     dat = pd.read_csv(
         fname,
-        delimiter=",",
+        delimiter=r"\s*,\s*",
+        engine="python",
         usecols=[ind for _field, ind in list_field_ind],
         skiprows=2,
         header=None,
@@ -360,7 +361,7 @@ def _read_part_csv_file_descriptor(fname: Union[str, "os.PathLike[str]"]):
 
     # Fields name from the default csv RAMSES sink algorithm in the yt default convention
     mapping = {
-        " # id": "particle_identifier",
+        "# id": "particle_identifier",
         "msink": "particle_mass",
         "x": "particle_position_x",
         "y": "particle_position_y",
@@ -386,11 +387,12 @@ def _read_part_csv_file_descriptor(fname: Union[str, "os.PathLike[str]"]):
     }
 
     # read the all file to get the number of particle
-    dat = pd.read_csv(fname, delimiter=",")
+    dat = pd.read_csv(fname, delimiter=r"\s*,\s*", engine="python")
     fields = []
     local_particle_count = len(dat)
 
     for varname in dat.columns:
+        varname = varname.strip()
         if varname in mapping:
             varname = mapping[varname]
         else: