Recent focus on the array of phenotypes expressed under differing environmental conditions, or phenotypic plasticity, has led to increased understanding of its genetic basis as well as its adaptive significance. However, the quantification of plasticity has proven difficult, hampered by both the limited number of environments over which plasticity may typically be assessed and by the need to assume, a priori, the general form of reaction norms under study. Our understanding of the shapes of continuous norms of reaction and, consequently, the subtle differences that may exist in shapes among genotypes or populations is rudimentary. Here, we propose the use of the loess smoothing function to analyze complex norms of reaction and to quantify total plasticity over many environments. A thermogradient incubator offers an ideal means to provide many environments for a demonstration of the use of the loess method. We test seed germination in three populations of two monocarpic plant species for population differentiation in plasticity to temperature. First, we test for differentiation in norms of reaction to 30 temperature environments among three populations of the monocarpic perennial, Lobelia inflata. The second demonstration assesses plasticity to eight temperature environments of three populations of the arctic‐alpine annual, Koenigia islandica. Our demonstration shows that the loess technique can detect significant genetic differentiation among populations in complex norms of reaction for both species studied, and suggests that the use of this procedure should be considered where the form of norms of reaction might be complex. The general applicability of the approach is discussed.