In a previous paper (6)* SAE 850178, the authors pointed out that the single-cycle gas simulation rig which they had developed would prove to be an invaluable experimental tool for the development of two-stroke cycle engine cylinders to attain better scavenging and trapping efficiency of the fresh charge. This paper reports on the use of that now proven experimental technique to examine one of the longest running, and hitherto unresolved, discussions in the field of small two-stroke cycle engines: is loop-scavenging really superior to cross-scavenging? All of the cross-scavenging tests in the paper are compared to tests conducted on loop-scavenged cylinders of the same basic geometry and which were reported previously to SAE. The main conclusion from the experimental investigation is that cross-scavenging is superior to loop-scavenging at low or modest scavenge ratios but is inferior at high scavenge ratios. However, another approach to cross-scavenging yields a design which retains and enhances the superiority at low or modest scavenge ratios and is at least equal to the very best loop-scavenging designs at the highest scavenge ratios. Equally important, this new cross-scavenging design lends itself to incorporation into a closely spaced multi-cylinder design without deterioration of those high scavenging and trapping efficiencies. The conclusions to be drawn from the tests on the new cross-scavenging design are that a potential exists for improving power and specific fuel consumption by 25% over the best loop-scavenging design.