diff --git a/test/runtests.jl b/test/runtests.jl
index 0b299eb1..acb91211 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -22,4 +22,5 @@ test_df = DataFrame(
 
 @testset "TidierData" verbose = true begin
     include("test_pivots.jl")
+    include("test_slice.jl")
 end
diff --git a/test/test_slice.jl b/test/test_slice.jl
new file mode 100644
index 00000000..fd778929
--- /dev/null
+++ b/test/test_slice.jl
@@ -0,0 +1,101 @@
+@testset "@slice()" verbose = true begin
+
+@testset "empty slice drops all rows" begin
+    df = DataFrame(g = [1, 1, 1], x = 1:3)
+    gdf = @group_by(df, g)
+
+    empty_df = DataFrame(g = Int[], x = Int[])
+    empty_gdf = @group_by(empty_df, g)
+
+    @test isequal(@slice(df), empty_df)
+    @test isequal(@slice(gdf), empty_gdf)
+end
+
+@testset "slicing DataFrame yields DataFrame" begin
+    df = DataFrame(x = 1:3)
+    @test isequal(@slice(df, 1), DataFrame(x = 1))
+end
+
+@testset "slice keeps positive indices, ignoring out of range" begin
+    gf = @group_by(DataFrame(g = [1, 2, 2, 3, 3, 3], id = 1:6), g)
+
+    #=
+    grouped dataframes don't behave exactly the same in Julia,
+    so you can't directly index into them like you can in R.
+    this feels like a close enough approximation of this test though.
+    =#
+    out = @slice(gf, 1)
+    @test isequal(@ungroup(out)[!, :id], [1, 2, 4])
+
+    out = @slice(gf, 2)
+    @test isequal(@ungroup(out)[!, :id], [3, 5])
+end
+
+@testset "slice keeps negative indices, ignoring out of range" begin
+    gf = @group_by(DataFrame(g = [1, 2, 2, 3, 3, 3], id = 1:6), g)
+
+    out = @slice(gf, -1)
+    @test isequal(@ungroup(out)[!, :id], [3, 5, 6])
+
+    out = @slice(gf, -(1:2))
+    @test isequal(@ungroup(out)[!, :id], 6)
+end
+
+@testset "slice errors if positive and negative indices mixed" begin
+    @test_throws "@slice() indices must either be all positive or all negative." @slice(DataFrame(), 1, -1)
+end
+
+@testset "slice errors if index is not numeric" begin
+    caught_error = false
+
+    try
+        eval(quote
+            df = DataFrame()
+            @slice(df, "a")
+        end)
+    catch e
+        caught_error = true
+        @test isa(e, MethodError) || isa(e, LoadError)
+    end
+
+    @test caught_error
+end
+
+@testset "slice keeps zero length groups" begin
+    df = DataFrame(
+        e = 1,
+        f = categorical([1, 1, 2, 2]; ordered = true, levels = 1:3),
+        g = [1, 1, 2, 2],
+        x = [1, 2, 1, 4]
+    )
+    df = @group_by(df, e, f, g)
+
+    @test isequal(
+        combine(@slice(df, 1), nrow => :size)[!, :size],
+        [1, 1, 0]
+    )
+end
+
+@testset "slice retains labels for zero length groups" begin
+    df = DataFrame(
+        e = 1,
+        f = categorical([1, 1, 2, 2]; ordered = true, levels = 1:3),
+        g = [1, 1, 2, 2],
+        x = [1, 2, 1, 4]
+    )
+    df = @group_by(df, e, f, g)
+
+    @test isequal(
+        @ungroup(@count(@slice(df, 1))),
+        DataFrame(
+            e = 1,
+            f = categorical(1:3; ordered = true, levels = 1:3),
+            g = [1, 2, missing],
+            n = [1, 1, 0]
+        )
+    )
+end
+end
+
+#TODO: slice_max tests
+#TODO: slice_min tests