diff --git a/src/protobuf.zig b/src/protobuf.zig
index 093b449..7c70298 100644
--- a/src/protobuf.zig
+++ b/src/protobuf.zig
@@ -723,7 +723,7 @@ fn VarintDecoderIterator(comptime T: type, comptime varint_type: VarintType) typ
             if (self.current_index < self.input.len) {
                 const raw_value = try decode_varint(u64, self.input[self.current_index..]);
                 defer self.current_index += raw_value.size;
-                return decode_varint_value(T, varint_type, raw_value.value);
+                return try decode_varint_value(T, varint_type, raw_value.value);
             }
             return null;
         }
@@ -808,7 +808,7 @@ pub const WireDecoderIterator = struct {
 };
 
 /// Get a real varint of type T from a raw u64 data.
-fn decode_varint_value(comptime T: type, comptime varint_type: VarintType, raw: u64) T {
+fn decode_varint_value(comptime T: type, comptime varint_type: VarintType, raw: u64) DecodingError!T {
     return switch (varint_type) {
         .ZigZagOptimized => switch (@typeInfo(T)) {
             .Int => {
@@ -825,7 +825,10 @@ fn decode_varint_value(comptime T: type, comptime varint_type: VarintType, raw:
                 else => @compileError("Invalid type " ++ @typeName(T) ++ " passed"),
             },
             .Bool => raw != 0,
-            .Enum => @as(T, @enumFromInt(@as(i32, @intCast(raw)))),
+            .Enum => if (raw > std.math.maxInt(u32))
+                error.InvalidInput
+            else
+                @as(T, @enumFromInt(@as(i32, @bitCast(@as(u32, @intCast(raw)))))),
             else => @compileError("Invalid type " ++ @typeName(T) ++ " passed"),
         },
     };
@@ -877,7 +880,7 @@ fn decode_packed_list(slice: []const u8, comptime list_type: ListType, comptime
 fn decode_value(comptime decoded_type: type, comptime ftype: FieldType, extracted_data: Extracted, allocator: Allocator) !decoded_type {
     return switch (ftype) {
         .Varint => |varint_type| switch (extracted_data.data) {
-            .RawValue => |value| decode_varint_value(decoded_type, varint_type, value),
+            .RawValue => |value| try decode_varint_value(decoded_type, varint_type, value),
             else => error.InvalidInput,
         },
         .FixedInt => switch (extracted_data.data) {
@@ -917,7 +920,7 @@ fn decode_data(comptime T: type, comptime field_desc: FieldDescriptor, comptime
             switch (list_type) {
                 .Varint => |varint_type| {
                     switch (extracted_data.data) {
-                        .RawValue => |value| try @field(result, field.name).append(decode_varint_value(child_type, varint_type, value)),
+                        .RawValue => |value| try @field(result, field.name).append(try decode_varint_value(child_type, varint_type, value)),
                         .Slice => |slice| try decode_packed_list(slice, list_type, child_type, &@field(result, field.name), allocator),
                     }
                 },
@@ -1724,12 +1727,12 @@ test "zigzag i32 - encode" {
 }
 
 test "zigzag i32/i64 - decode" {
-    try testing.expectEqual(@as(i32, 1), decode_varint_value(i32, .ZigZagOptimized, 2));
-    try testing.expectEqual(@as(i32, -2), decode_varint_value(i32, .ZigZagOptimized, 3));
-    try testing.expectEqual(@as(i32, -500), decode_varint_value(i32, .ZigZagOptimized, 999));
-    try testing.expectEqual(@as(i64, -500), decode_varint_value(i64, .ZigZagOptimized, 999));
-    try testing.expectEqual(@as(i64, -500), decode_varint_value(i64, .ZigZagOptimized, 999));
-    try testing.expectEqual(@as(i64, -0x80000000), decode_varint_value(i64, .ZigZagOptimized, 0xffffffff));
+    try testing.expectEqual(@as(i32, 1), try decode_varint_value(i32, .ZigZagOptimized, 2));
+    try testing.expectEqual(@as(i32, -2), try decode_varint_value(i32, .ZigZagOptimized, 3));
+    try testing.expectEqual(@as(i32, -500), try decode_varint_value(i32, .ZigZagOptimized, 999));
+    try testing.expectEqual(@as(i64, -500), try decode_varint_value(i64, .ZigZagOptimized, 999));
+    try testing.expectEqual(@as(i64, -500), try decode_varint_value(i64, .ZigZagOptimized, 999));
+    try testing.expectEqual(@as(i64, -0x80000000), try decode_varint_value(i64, .ZigZagOptimized, 0xffffffff));
 }
 
 test "zigzag i64 - encode" {
@@ -1750,3 +1753,29 @@ test "incorrect data - decode" {
 
     try testing.expectError(error.InvalidInput, value);
 }
+
+test "incorrect data - simple varint" {
+    // Incorrectly serialized protobufs can place a varint with a decoded value
+    // greater than std.math.maxInt(u32) into the slot an enum is supposed to
+    // fill. Since this library represents a decoded varint as a u64 -- the max
+    // possible valid varint width -- that data can make its way deep into the
+    // decode_varint_value routine. This test checks that we handle such failures
+    // gracefully rather than panicking.
+    const max_u64 = decode_varint_value(enum(i32) { a, b, c }, .Simple, (1 << 64) - 1);
+    const barely_too_big = decode_varint_value(enum(i32) { a, b, c }, .Simple, 1 << 32);
+
+    try std.testing.expectError(error.InvalidInput, max_u64);
+    try std.testing.expectError(error.InvalidInput, barely_too_big);
+}
+
+test "correct data - simple varint" {
+    const enum_a = try decode_varint_value(enum(i32) { a = -1, b = 0, c = 1, d = 2 }, .Simple, (1 << 32) - 1);
+    const enum_b = try decode_varint_value(enum(i32) { a = -1, b = 0, c = 1, d = 2 }, .Simple, 0);
+    const enum_c = try decode_varint_value(enum(i32) { a = -1, b = 0, c = 1, d = 2 }, .Simple, 1);
+    const enum_d = try decode_varint_value(enum(i32) { a = -1, b = 0, c = 1, d = 2 }, .Simple, 2);
+
+    try std.testing.expectEqual(.a, enum_a);
+    try std.testing.expectEqual(.b, enum_b);
+    try std.testing.expectEqual(.c, enum_c);
+    try std.testing.expectEqual(.d, enum_d);
+}