diff --git a/ext/IntegralsMCIntegrationExt.jl b/ext/IntegralsMCIntegrationExt.jl
index e14233d..c761ff1 100644
--- a/ext/IntegralsMCIntegrationExt.jl
+++ b/ext/IntegralsMCIntegrationExt.jl
@@ -3,10 +3,10 @@ module IntegralsMCIntegrationExt
 using MCIntegration, Integrals
 
 function Integrals.__solvebp_call(prob::IntegralProblem, alg::VEGASMC, sensealg, domain, p;
-    reltol = nothing, abstol = nothing, maxiters = 1000)
+    reltol = nothing, abstol = nothing, maxiters = 10)
     lb, ub = domain
     mid = vec(collect((lb + ub) / 2))
-    vars = Continuous(vec([tuple(a,b) for (a,b) in zip(lb, ub)]))
+    var = Continuous(vec([tuple(a,b) for (a,b) in zip(lb, ub)]))
 
     if prob.f isa BatchIntegralFunction
         error("VEGASMC doesn't support batching. See https://github.com/numericalEFT/MCIntegration.jl/issues/29")
@@ -33,9 +33,8 @@ function Integrals.__solvebp_call(prob::IntegralProblem, alg::VEGASMC, sensealg,
             end
         end
         dof = ones(Int, length(f0)) # each composite Continuous var gets 1 dof
-        res = integrate(f_, inplace=isinplace(prob), var=vars, dof=dof, solver=:vegasmc,
-            neval=alg.neval, niter=min(maxiters,alg.niter), block=alg.block, adapt=alg.adapt,
-            gamma=alg.gamma, verbose=alg.verbose, debug=alg.debug, type=eltype(f0), print=-2)
+        res = integrate(f_; var, dof, inplace=isinplace(prob), type=eltype(f0),
+            solver=:vegasmc, niter=maxiters, verbose=-2,  print=-2, alg.kws...)
         out, err, chi = if f0 isa Number
             map(only, (res.mean, res.stdev, res.chi2))
         else
diff --git a/src/algorithms.jl b/src/algorithms.jl
index 972c464..f2c1dfc 100644
--- a/src/algorithms.jl
+++ b/src/algorithms.jl
@@ -372,18 +372,16 @@ function ArblibJL(; check_analytic=false, take_prec=false, warn_on_no_convergenc
 end
 
 """
-    VEGASMC(; neval=10^4, niter=20, block=16, adapt=true, gamma=1.0, verbose=-2, debug=false)
+    VEGASMC(; kws...)
 
 Markov-chain based Vegas algorithm from MCIntegration.jl
+
+Refer to
+[`MCIntegration.integrate`](https://numericaleft.github.io/MCIntegration.jl/dev/lib/montecarlo/#MCIntegration.integrate-Tuple{Function})
+for documentation on the keywords, which are passed directly to the solver with a set of
+defaults that works for conforming integrands.
 """
-struct VEGASMC <: SciMLBase.AbstractIntegralAlgorithm
-    neval::Int
-    niter::Int
-    block::Int
-    adapt::Bool
-    gamma::Float64
-    verbose::Int
-    debug::Bool
+struct VEGASMC{K<:NamedTuple} <: SciMLBase.AbstractIntegralAlgorithm
+    kws::K
 end
-VEGASMC(; neval=10^4, niter=20, block=16, adapt=true, gamma=1.0, verbose=-2, debug=false) =
-    VEGASMC(neval, niter, block, adapt, gamma, verbose, debug)
+VEGASMC(; kws...) = VEGASMC(NamedTuple(kws))
diff --git a/test/interface_tests.jl b/test/interface_tests.jl
index 57457e2..1a0137c 100644
--- a/test/interface_tests.jl
+++ b/test/interface_tests.jl
@@ -18,7 +18,7 @@ alg_req = Dict(
         max_dim = Inf, allows_iip = true),
     # VEGAS() => (nout = 1, allows_batch = true, min_dim = 2, max_dim = Inf,
     #     allows_iip = true),
-    VEGASMC() => (nout = Inf, allows_batch = false, min_dim = 1, max_dim = Inf,
+    VEGASMC(seed=42) => (nout = Inf, allows_batch = false, min_dim = 1, max_dim = Inf,
         allows_iip = true),
     CubatureJLh() => (nout = Inf, allows_batch = true, min_dim = 1,
         max_dim = Inf, allows_iip = true),