In this paper I examine the relationship between Pareto-optimality and group size in linear public goods games or experiments. In particular, I use the standard setting of homogeneous linear public goods experiments and apply a recently developed tool to identify all Pareto-optimal allocations in such settings. It turns out that under any conceivable circumstances, ceteris paribus, small groups have a higher Pareto-ratio (Pareto-optimal allocations over total allocations) than large groups. Hence, if Pareto-optimality of an allocation is a property that makes such allocations acceptable and maintainable, small groups will find is easier to provide Pareto-optimal amounts of a public good than large groups. This is a novel reasoning for Mancur Olson’s claim, in particular, with respect to what he has termed inclusive goods and inclusive groups.