diff --git a/Project.toml b/Project.toml
index 1bc53ff4..cac18ad2 100644
--- a/Project.toml
+++ b/Project.toml
@@ -1,7 +1,7 @@
 name = "MLJFlux"
 uuid = "094fc8d1-fd35-5302-93ea-dabda2abf845"
 authors = ["Anthony D. Blaom <anthony.blaom@gmail.com>", "Ayush Shridhar <ayush.shridhar1999@gmail.com>"]
-version = "0.3.0"
+version = "0.3.1"
 
 [deps]
 CategoricalArrays = "324d7699-5711-5eae-9e2f-1d82baa6b597"
diff --git a/src/mlj_model_interface.jl b/src/mlj_model_interface.jl
index f3af0774..5ffe903f 100644
--- a/src/mlj_model_interface.jl
+++ b/src/mlj_model_interface.jl
@@ -59,6 +59,7 @@ function MLJModelInterface.fit(model::MLJFluxModel,
         build(model, rng, shape) |> move
     catch ex
         @error ERR_BUILDER
+        rethrow()
     end
 
     penalty = Penalty(model)
diff --git a/test/mlj_model_interface.jl b/test/mlj_model_interface.jl
index 6f07a661..a2e6c8fe 100644
--- a/test/mlj_model_interface.jl
+++ b/test/mlj_model_interface.jl
@@ -52,3 +52,31 @@ end
     @test length(losses) == 10
 end
 
+mutable struct LisasBuilder
+  n1::Int
+end
+
+@testset "builder errors and issue #237" begin
+    # create a builder with an intentional flaw;
+    # `Chains` is undefined - it should be `Chain`
+    function MLJFlux.build(builder::LisasBuilder, rng, nin, nout)
+        return Flux.Chains(
+            Flux.Dense(nin, builder.n1),
+            Flux.Dense(builder.n1, nout)
+        )
+    end
+
+    model = NeuralNetworkRegressor(
+        epochs = 2,
+        batch_size = 32,
+        builder = LisasBuilder(10),
+    )
+
+    X, y = @load_boston
+    @test_logs(
+        (:error, MLJFlux.ERR_BUILDER),
+        @test_throws UndefVarError(:Chains) MLJBase.fit(model, 0, X, y)
+    )
+end
+
+true