diff --git a/docs/RUGNUX_TUTORIAL.md b/docs/RUGNUX_TUTORIAL.md
index 56c6e0977..bfdb25d02 100644
--- a/docs/RUGNUX_TUTORIAL.md
+++ b/docs/RUGNUX_TUTORIAL.md
@@ -138,7 +138,9 @@ into the merged files.
 Run **fully de novo** (no `-C`/`-S`) for the best result — supplying a cell or space group up front
 tends to *degrade* low-symmetry cases. A `-S` group whose Bravais lattice the crystal turns out not to
 have stops the run and names the cell that was indexed, rather than merging in a frame the reflections
-are not in; where the lattice does have that group's setting, the reflections are reindexed into it. `--scaling-high-resolution` (set it to your expected
+are not in; where the lattice does have that group's setting, the reflections are reindexed into it.
+A primitive `-S` group on a lattice indexed as centred is the exception: it is merged as asked, with a
+warning. `--scaling-high-resolution` (set it to your expected
 resolution) sharpens both the space-group search and the error model. To tune the first pass use
 `--two-pass-rotation=100` (or `-R100` — the first-pass image count); to force the sweep to be
 treated as independent stills use `--force-still`.
diff --git a/rugnux/RugnuxScaleMerge.cpp b/rugnux/RugnuxScaleMerge.cpp
index 6478ebcd9..dbf89d206 100644
--- a/rugnux/RugnuxScaleMerge.cpp
+++ b/rugnux/RugnuxScaleMerge.cpp
@@ -1223,8 +1223,26 @@ bool Rugnux::ScaleMergeAndSymmetry(PipelineLocals &p) {
     // question that was not asked. Name the cell that WAS indexed so the user can act on it.
     if (user_fixed_sg_ && end_msg.rotation_lattice_type.has_value() && end_msg.unit_cell.has_value()) {
         const gemmi::SpaceGroup &fixed_sg = *user_fixed_sg_;
-        const bool centring_wrong =
-            end_msg.rotation_lattice_type->centering != fixed_sg.centring_type();
+        // A PRIMITIVE group on a lattice indexed as centred is not that conflict: the reflections are
+        // integrated on the conventional cell, so the group's frame is theirs and its arithmetic holds;
+        // it only keeps the reflections the centring would extinguish. Whether those are absent is the
+        // data's to say, and the user chose the group - so warn, and merge in it.
+        const bool primitive_on_centred = fixed_sg.centring_type() == 'P'
+                                          && end_msg.rotation_lattice_type->centering != 'P';
+        if (primitive_on_centred) {
+            const auto &uc = *end_msg.unit_cell;
+            logger.Warning("The space group {} was fixed for this run and is primitive, but the indexing "
+                           "found a {}-centred {} lattice (a={:.3f} b={:.3f} c={:.3f} alpha={:.2f} "
+                           "beta={:.2f} gamma={:.2f}): merging in {} as asked. If the centring is real, "
+                           "the reflections it extinguishes merge to noise; if they are only weak, a "
+                           "pseudo-translation is the likely reason. Without -S the space-group search "
+                           "decides it from the intensities",
+                           fixed_sg.xhm(), end_msg.rotation_lattice_type->centering,
+                           gemmi::crystal_system_str(end_msg.rotation_lattice_type->crystal_system),
+                           uc.a, uc.b, uc.c, uc.alpha, uc.beta, uc.gamma, fixed_sg.xhm());
+        }
+        const bool centring_wrong = !primitive_on_centred
+            && end_msg.rotation_lattice_type->centering != fixed_sg.centring_type();
         // The same two tests as the re-seating above, asked again now that it has had its chance.
         const double violation = MetricViolation(*end_msg.unit_cell, fixed_sg);
         if (centring_wrong || violation > MAX_METRIC_VIOLATION) {