diff --git a/augur/translate.py b/augur/translate.py
index 4e19f61c7..485266de7 100644
--- a/augur/translate.py
+++ b/augur/translate.py
@@ -328,8 +328,7 @@ def sequences_vcf(reference_fasta, vcf):
     [0] The sequences as a dict of dicts. sequences → <NODE_NAME> → <POS> → <ALT_NUC> where <POS> is a 0-based int
     [1] The sequence of the provided `reference_fasta` (string)
     """
-    if not reference_fasta:
-        raise AugurError("A reference Fasta is required with VCF-format input")
+    assert reference_fasta is not None
     compress_seq = read_vcf(vcf, reference_fasta)
     sequences = compress_seq['sequences']
     ref = compress_seq['reference']
@@ -385,8 +384,25 @@ def check_arg_combinations(args, is_vcf):
     This checking shouldn't be used by downstream code to assume arguments exist, however by checking for
     invalid combinations up-front we can exit quickly.
     """
-    if not is_vcf and (args.vcf_reference or args.vcf_reference_output):
-        raise AugurError("Arguments '--vcf-reference' and/or '--vcf-reference-output' are only applicable if the input ('--ancestral-sequences') is VCF")
+
+    if is_vcf:
+        if not args.vcf_reference:
+            raise AugurError("A reference FASTA (--vcf-reference) is required with VCF-format input")
+    else:
+        if args.vcf_reference or args.vcf_reference_output:
+            raise AugurError("Arguments '--vcf-reference' and/or '--vcf-reference-output' are only applicable if the input ('--ancestral-sequences') is VCF")
+    
+    if args.alignment_output:
+        if is_vcf:
+            if not is_filename_vcf(args.alignment_output):
+                raise AugurError("When using a VCF input the --alignment-output filename must also be a VCF file")
+            if not args.vcf_reference_output:
+                raise AugurError("When using a VCF input and --alignment-output, we now require you to specify the --vcf-reference-output as well")
+        else:
+            if is_filename_vcf(args.alignment_output):
+                raise AugurError("When using a non-VCF input the --alignment-output filename must not be a VCF file")
+    if args.vcf_reference_output and not args.alignment_output:
+        raise AugurError("The VCF reference output (--vcf-reference-output) needs --alignment-output")
 
 
 def run(args):
@@ -474,15 +490,12 @@ def run(args):
     write_json(output_data, out_name)
     print("amino acid mutations written to", out_name, file=sys.stdout)
 
-    ## write alignments to file is requested
+    ## write alignments to file if requested
     if args.alignment_output:
         if is_vcf:
-            ## write VCF-style output if requested
-            fileEndings = -1
-            if args.alignment_output.lower().endswith('.gz'):
-                fileEndings = -2
-            vcf_out_ref = args.vcf_reference_output or '.'.join(args.alignment_output.split('.')[:fileEndings]) + '_reference.fasta'
-            write_VCF_translation(translations, args.alignment_output, vcf_out_ref)
+            assert is_filename_vcf(args.alignment_output)
+            assert args.vcf_reference_output is not None
+            write_VCF_translation(translations, args.alignment_output, args.vcf_reference_output)
         else:
             ## write fasta-style output if requested
             if '%GENE' in args.alignment_output:
diff --git a/tests/functional/translate/cram/invalid-args.t b/tests/functional/translate/cram/invalid-args.t
new file mode 100644
index 000000000..e54e64601
--- /dev/null
+++ b/tests/functional/translate/cram/invalid-args.t
@@ -0,0 +1,57 @@
+Setup
+
+  $ export AUGUR="${AUGUR:-$TESTDIR/../../../../bin/augur}"
+  $ export SCRIPTS="$TESTDIR/../../../../scripts"
+  $ export ANC_DATA="$TESTDIR/../../ancestral/data/simple-genome"
+  $ export DATA="$TESTDIR/../data/simple-genome"
+
+
+Input JSON + VCF output is not possible (and vice-versa)
+
+  $ ${AUGUR} translate \
+  >  --tree $ANC_DATA/tree.nwk \
+  >  --ancestral-sequences $ANC_DATA/nt_muts.ref-seq.json \
+  >  --reference-sequence $DATA/reference.gff \
+  >  --alignment-output "translations.vcf"
+  ERROR: When using a non-VCF input the --alignment-output filename must not be a VCF file
+  [2]
+
+  $ ${AUGUR} translate \
+  >  --tree $ANC_DATA/tree.nwk \
+  >  --ancestral-sequences input.vcf \
+  >  --vcf-reference reference_in.fasta \
+  >  --reference-sequence $DATA/reference.gff \
+  >  --alignment-output "translations.fasta"
+  ERROR: When using a VCF input the --alignment-output filename must also be a VCF file
+  [2]
+
+The arg --vcf-reference-output needs --alignment-output
+
+  $ ${AUGUR} translate \
+  >  --tree $ANC_DATA/tree.nwk \
+  >  --ancestral-sequences input.vcf \
+  >  --reference-sequence $DATA/reference.gff \
+  >  --vcf-reference reference_in.fasta \
+  >  --vcf-reference-output reference_out.fasta
+  ERROR: The VCF reference output (--vcf-reference-output) needs --alignment-output
+  [2]
+
+If VCF input with requested alignment output then we require --vcf-reference-output
+(This was not the case in augur 23.1.1 and earlier, where we would automatically create a filename)
+  $ ${AUGUR} translate \
+  >  --tree $ANC_DATA/tree.nwk \
+  >  --ancestral-sequences input.vcf \
+  >  --vcf-reference reference_in.fasta \
+  >  --reference-sequence $DATA/reference.gff \
+  >  --alignment-output "translations.vcf"
+  ERROR: When using a VCF input and --alignment-output, we now require you to specify the --vcf-reference-output as well
+  [2]
+
+VCF input must have a FASTA reference provided
+  $ ${AUGUR} translate \
+  >  --tree $ANC_DATA/tree.nwk \
+  >  --ancestral-sequences input.vcf \
+  >  --reference-sequence $DATA/reference.gff \
+  >  --alignment-output "translations.vcf"
+  ERROR: A reference FASTA (--vcf-reference) is required with VCF-format input
+  [2]