The effects of natural selection on quantitative traits were compared in 17 single hybrid barley populations which had been developed either by growing the bulk material continuously in each of two locations, Macdonald College and La Pocatière, or by alternating between these locations annually or after 2–8 yr per location. Comparisons were made by testing the F10 and F15 generations of each of the 17 populations at the two locations for 2 yr. Natural selection favored longer awns, shorter spikes, smaller flag leaves, fewer spikes per plot, and, perhaps, heavier grains. Natural selection had very little effect, however, on the number of grains per spike or grain yield. Variation among populations was observed only for awn length, spike length, flag leaf area, and number of grains per spike in the F15 generation. This indicates that the rate of change for the four traits was affected by the propagation environments but was not affected by the frequency of alternation because no distinct pattern was found among these populations. These populations in both generations responded similarly to the testing environments except for awn length in F15. It is suggested that wider diversity of locality is needed in order to evaluate effectively the usefulness of a natural selection procedure involving the alternating of a segregating population.