diff --git a/docs/src/apimanual.md b/docs/src/apimanual.md
index 15c0818617..68ebde00a7 100644
--- a/docs/src/apimanual.md
+++ b/docs/src/apimanual.md
@@ -1084,3 +1084,46 @@ const cache = MOIU.UniversalFallback(ModelData{Float64}())
 const cached = MOIU.CachingOptimizer(cache, optimizer)
 const bridged = MOIB.full_bridge_optimizer(cached, Float64)
 ```
+
+### Benchmarking
+
+To aid the development of efficient solver wrappers, MathOptInterface provides
+benchmarking functionality. Benchmarking a wrapper follows a two-step process.
+
+First, prior to making changes, run and save the benchmark results on a given
+benchmark suite as follows:
+
+```julia
+using SolverPackage, MathOptInterface
+
+const MOI = MathOptInterface
+
+suite = MOI.Benchmarks.suite() do
+    SolverPackage.Optimizer()
+end
+
+MOI.Benchmarks.create_baseline(
+    suite, "current"; directory = "/tmp", verbose = true
+)
+```
+Use the `exclude` argument to [`Benchmarks.suite`](@ref) to
+exclude benchmarks that the solver doesn't support.
+
+Second, after making changes to the package, re-run the benchmark suite and
+compare to the prior saved results:
+
+```julia
+using SolverPackage, MathOptInterface
+
+const MOI = MathOptInterface
+
+suite = MOI.Benchmarks.suite() do
+    SolverPackage.Optimizer()
+end
+
+MOI.Benchmarks.compare_against_baseline(
+    suite, "current"; directory = "/tmp", verbose = true
+)
+```
+
+This comparison will create a report detailing improvements and regressions.
diff --git a/docs/src/apireference.md b/docs/src/apireference.md
index 15837b3093..650f19cc7d 100644
--- a/docs/src/apireference.md
+++ b/docs/src/apireference.md
@@ -546,3 +546,14 @@ Utilities.load_variables
 Utilities.load
 Utilities.load_constraint
 ```
+
+## Benchmarks
+
+Functions to help benchmark the performance of solver wrappers. See
+[Benchmarking](@ref) for more details.
+
+```@docs
+Benchmarks.suite
+Benchmarks.create_baseline
+Benchmarks.compare_against_baseline
+```
diff --git a/src/Benchmarks/Benchmarks.jl b/src/Benchmarks/Benchmarks.jl
new file mode 100644
index 0000000000..b1b607417c
--- /dev/null
+++ b/src/Benchmarks/Benchmarks.jl
@@ -0,0 +1,236 @@
+module Benchmarks
+
+using BenchmarkTools, MathOptInterface
+
+const MOI = MathOptInterface
+const BENCHMARKS = Dict{String, Function}()
+
+"""
+    suite(
+        new_model::Function;
+        exclude::Vector{Regex} = Regex[]
+    )
+
+Create a suite of benchmarks. `new_model` should be a function that takes no
+arguments, and returns a new instance of the optimizer you wish to benchmark.
+
+Use `exclude` to exclude a subset of benchmarks.
+
+### Examples
+
+```julia
+suite() do
+    GLPK.Optimizer()
+end
+
+suite(exclude = [r"delete"]) do
+    Gurobi.Optimizer(OutputFlag=0)
+end
+````
+"""
+function suite(new_model::Function; exclude::Vector{Regex} = Regex[])
+    group = BenchmarkGroup()
+    for (name, func) in BENCHMARKS
+        any(occursin.(exclude, Ref(name))) && continue
+        group[name] = @benchmarkable $func($new_model)
+    end
+    return group
+end
+
+"""
+    create_baseline(suite, name::String; directory::String = ""; kwargs...)
+
+Run all benchmarks in `suite` and save to files called `name` in `directory`.
+
+Extra `kwargs` are based to `BenchmarkTools.run`.
+
+### Examples
+
+```julia
+my_suite = suite(() -> GLPK.Optimizer())
+create_baseline(my_suite, "glpk_master"; directory = "/tmp", verbose = true)
+```
+"""
+function create_baseline(
+    suite::BenchmarkTools.BenchmarkGroup, name::String; directory::String = "",
+    kwargs...
+)
+    tune!(suite)
+    BenchmarkTools.save(joinpath(directory, name * "_params.json"), params(suite))
+    results = run(suite; kwargs...)
+    BenchmarkTools.save(joinpath(directory, name * "_baseline.json"), results)
+    return
+end
+
+"""
+    compare_against_baseline(
+        suite, name::String; directory::String = "",
+        report_filename::String = "report.txt"
+    )
+
+Run all benchmarks in `suite` and compare against files called `name` in
+`directory` that were created by a call to `create_baseline`.
+
+A report summarizing the comparison is written to `report_filename` in
+`directory`.
+
+Extra `kwargs` are based to `BenchmarkTools.run`.
+
+### Examples
+
+```julia
+my_suite = suite(() -> GLPK.Optimizer())
+compare_against_baseline(
+    my_suite, "glpk_master"; directory = "/tmp", verbose = true
+)
+```
+"""
+function compare_against_baseline(
+    suite::BenchmarkTools.BenchmarkGroup, name::String;
+    directory::String = "", report_filename::String = "report.txt", kwargs...
+)
+    params_filename = joinpath(directory, name * "_params.json")
+    baseline_filename = joinpath(directory, name * "_baseline.json")
+    if !isfile(params_filename) || !isfile(baseline_filename)
+        error("You create a baseline with `create_baseline` first.")
+    end
+    loadparams!(
+        suite, BenchmarkTools.load(params_filename)[1], :evals, :samples
+    )
+    new_results = run(suite; kwargs...)
+    old_results = BenchmarkTools.load(baseline_filename)[1]
+    open(joinpath(directory, report_filename), "w") do io
+        println(stdout, "\n========== Results ==========")
+        println(io,     "\n========== Results ==========")
+        for key in keys(new_results)
+            judgement = judge(
+                BenchmarkTools.median(new_results[key]),
+                BenchmarkTools.median(old_results[key])
+            )
+            println(stdout, "\n", key)
+            println(io,     "\n", key)
+            show(stdout, MIME"text/plain"(), judgement)
+            show(io, MIME"text/plain"(), judgement)
+        end
+    end
+    return
+end
+
+###
+### Benchmarks
+###
+
+macro add_benchmark(f)
+    name = f.args[1].args[1]
+    return quote
+        $(esc(f))
+        BENCHMARKS[String($(Base.Meta.quot(name)))] = $(esc(name))
+    end
+end
+
+@add_benchmark function add_variable(new_model)
+    model = new_model()
+    for i in 1:10_000
+        MOI.add_variable(model)
+    end
+    return model
+end
+
+@add_benchmark function add_variables(new_model)
+    model = new_model()
+    MOI.add_variables(model, 10_000)
+    return model
+end
+
+@add_benchmark function add_variable_constraint(new_model)
+    model = new_model()
+    x = MOI.add_variables(model, 10_000)
+    for (i, xi) in enumerate(x)
+        MOI.add_constraint(model, MOI.SingleVariable(xi), MOI.LessThan(1.0 * i))
+    end
+    return model
+end
+
+@add_benchmark function add_variable_constraints(new_model)
+    model = new_model()
+    x = MOI.add_variables(model, 10_000)
+    MOI.add_constraints(
+        model,
+        MOI.SingleVariable.(x),
+        MOI.LessThan.(1.0:10_000.0)
+    )
+    return model
+end
+
+@add_benchmark function delete_variable(new_model)
+    model = new_model()
+    x = MOI.add_variables(model, 1_000)
+    MOI.add_constraint.(model, MOI.SingleVariable.(x), Ref(MOI.LessThan(1.0)))
+    MOI.delete.(model, x)
+    return model
+end
+
+@add_benchmark function delete_variable_constraint(new_model)
+    model = new_model()
+    x = MOI.add_variables(model, 1_000)
+    cons = MOI.add_constraint.(model, MOI.SingleVariable.(x), Ref(MOI.LessThan(1.0)))
+    for con in cons
+        MOI.delete(model, con)
+    end
+    cons = MOI.add_constraint.(model, MOI.SingleVariable.(x), Ref(MOI.LessThan(1.0)))
+    MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE)
+    MOI.set(model,
+        MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+        MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(1.0, x), 0.0)
+    )
+    MOI.optimize!(model)
+    for con in cons
+        MOI.delete(model, con)
+    end
+    return model
+end
+
+@add_benchmark function add_constraint(new_model)
+    model = new_model()
+    index = MOI.add_variables(model, 10_000)
+    for (i, x) in enumerate(index)
+        MOI.add_constraint(
+            model,
+            MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+            MOI.LessThan(1.0 * i)
+        )
+    end
+    return model
+end
+
+@add_benchmark function add_constraints(new_model)
+    model = new_model()
+    x = MOI.add_variables(model, 10_000)
+    MOI.add_constraints(
+        model,
+        [MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, xi)], 0.0) for xi in x],
+        MOI.LessThan.(1:1.0:10_000)
+    )
+    return model
+end
+
+@add_benchmark function delete_constraint(new_model)
+    model = new_model()
+    index = MOI.add_variables(model, 1_000)
+    cons = Vector{
+        MOI.ConstraintIndex{MOI.ScalarAffineFunction{Float64}, MOI.LessThan{Float64}}
+    }(undef, 1_000)
+    for (i, x) in enumerate(index)
+        cons[i] = MOI.add_constraint(
+            model,
+            MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+            MOI.LessThan(1.0 * i)
+        )
+    end
+    for con in cons
+        MOI.delete(model, con)
+    end
+    return model
+end
+
+end
diff --git a/src/MathOptInterface.jl b/src/MathOptInterface.jl
index e4838c17d5..f71a920c87 100644
--- a/src/MathOptInterface.jl
+++ b/src/MathOptInterface.jl
@@ -121,5 +121,6 @@ include("nlp.jl")
 include("Utilities/Utilities.jl") # MOI.Utilities
 include("Test/Test.jl")           # MOI.Test
 include("Bridges/Bridges.jl")     # MOI.Bridges
+include("Benchmarks/Benchmarks.jl")
 
 end
diff --git a/test/Benchmarks/Benchmarks.jl b/test/Benchmarks/Benchmarks.jl
new file mode 100644
index 0000000000..b391376b46
--- /dev/null
+++ b/test/Benchmarks/Benchmarks.jl
@@ -0,0 +1,60 @@
+using MathOptInterface, Test
+
+const MOI = MathOptInterface
+const MOIU = MOI.Utilities
+
+MOIU.@model(
+    BenchmarkModel,
+    (), (MOI.LessThan, ), (), (),
+    (), (MOI.ScalarAffineFunction, ), (), ()
+)
+
+const NUM_BENCHMARKS = length(MOI.Benchmarks.BENCHMARKS)
+
+@testset "suite" begin
+    suite = MOI.Benchmarks.suite() do
+        MOIU.MockOptimizer(BenchmarkModel{Float64}())
+    end
+    @test length(suite.data) == NUM_BENCHMARKS
+
+    suite = MOI.Benchmarks.suite(
+        exclude = [r"delete_"]
+    ) do
+        MOIU.MockOptimizer(BenchmarkModel{Float64}())
+    end
+    # Note: update this value whenever more benchmarks are added to
+    # `src/Benchmarks/Benchmarks.jl`.
+    @test 6 <= length(suite.data) <= NUM_BENCHMARKS - 3
+end
+
+@testset "Perform benchmark" begin
+    params = joinpath(@__DIR__, "baseline_params.json")
+    baseline = joinpath(@__DIR__, "baseline_baseline.json")
+    @test !isfile(params)
+    @test !isfile(baseline)
+    @testset "create_baseline" begin
+        suite = MOI.Benchmarks.suite() do
+            MOIU.MockOptimizer(BenchmarkModel{Float64}())
+        end
+        MOI.Benchmarks.create_baseline(
+            suite, "baseline"; directory=@__DIR__, seconds = 2, verbose = true
+        )
+    end
+    @test isfile(params)
+    @test isfile(baseline)
+    @testset "compare_against_baseline" begin
+        suite = MOI.Benchmarks.suite() do
+            MOIU.MockOptimizer(BenchmarkModel{Float64}())
+        end
+        MOI.Benchmarks.compare_against_baseline(
+            suite, "baseline"; directory=@__DIR__, seconds = 2, verbose = true
+        )
+    end
+    rm(params)
+    rm(baseline)
+    @testset "Report" begin
+        report = read(joinpath(@__DIR__, "report.txt"), String)
+        @test occursin("=> invariant", report)
+    end
+    rm(joinpath(@__DIR__, "report.txt"))
+end
diff --git a/test/runtests.jl b/test/runtests.jl
index 13b56ac18f..959f56b74e 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -37,5 +37,9 @@ end
     include("Bridges/Bridges.jl")
 end
 
+@testset "MOI.Benchmarks" begin
+    include("Benchmarks/Benchmarks.jl")
+end
+
 # Test hygiene of @model macro
 include("hygiene.jl")