diff --git a/test/staticsize.jl b/test/staticsize.jl
index 6ff2b0f6..e0902f43 100644
--- a/test/staticsize.jl
+++ b/test/staticsize.jl
@@ -135,3 +135,71 @@ end
     @test sum2_10turbo(A) ≈ sum(A)
   end
 end
+
+# Test for Issue #543: W=1 nested VecUnroll store on ARM
+# This tests the case where vector width is 1 (scalar) with nested unrolling
+function issue543_noavx!(data_out, matrix, data_in)
+  for j in axes(data_out, 3), i in axes(data_out, 2), v in axes(data_out, 1)
+    res = zero(eltype(data_out))
+    for jj in axes(matrix, 2)
+      res += matrix[j, jj] * data_in[v, i, jj]
+    end
+    data_out[v, i, j] = res
+  end
+  return nothing
+end
+
+function issue543_turbo!(data_out, matrix, data_in)
+  @turbo for j in axes(data_out, 3), i in axes(data_out, 2), v in axes(data_out, 1)
+    res = zero(eltype(data_out))
+    for jj in axes(matrix, 2)
+      res += matrix[j, jj] * data_in[v, i, jj]
+    end
+    data_out[v, i, j] = res
+  end
+  return nothing
+end
+
+@testset "Issue #543: W=1 Nested VecUnroll" begin
+  # Test with static first dimension
+  for v in 1:4, n in 2:8
+    data_out_ref = StrideArray(undef, StaticInt(v), StaticInt(n), StaticInt(n))
+    data_out_turbo = StrideArray(undef, StaticInt(v), StaticInt(n), StaticInt(n))
+    matrix = StrideArray(undef, StaticInt(n), StaticInt(n))
+    data_in = rand(v, n, n)
+
+    matrix .= rand.()
+
+    fill!(data_out_ref, 0.0)
+    fill!(data_out_turbo, 0.0)
+
+    issue543_noavx!(data_out_ref, matrix, data_in)
+
+    # This is broken on Apple ARM CPUs (Apple M series) for some reason.
+    # TODO: Fix the underlying issue!
+    if (v == 1) && Sys.isapple() && Sys.ARCH == :aarch64
+      @test_skip issue543_turbo!(data_out_turbo, matrix, data_in)
+    else
+      @test_nowarn issue543_turbo!(data_out_turbo, matrix, data_in)
+      @test data_out_turbo ≈ data_out_ref
+    end
+  end
+
+  # Test with non-static first but static other dimensions
+  for v in 1:4, n in 2:8
+    data_out_ref = StrideArray(undef, v, StaticInt(n), StaticInt(n))
+    data_out_turbo = StrideArray(undef, v, StaticInt(n), StaticInt(n))
+    matrix = StrideArray(undef, StaticInt(n), StaticInt(n))
+    data_in = rand(v, n, n)
+
+    matrix .= rand.()
+
+    fill!(data_out_ref, 0.0)
+    fill!(data_out_turbo, 0.0)
+
+    issue543_noavx!(data_out_ref, matrix, data_in)
+    issue543_turbo!(data_out_turbo, matrix, data_in)
+
+    @test data_out_turbo ≈ data_out_ref
+  end
+end