I elaborate an hypothesis to explain inconsistent empirical findings comparing phenotypic plasticity in colonizing populations or species with plasticity from their native or ancestral range. Quantitative genetic theory on the evolution of plasticity reveals that colonization of a novel environment can cause a transient increase in plasticity: a rapid initial increase in plasticity accelerates evolution of a new optimal phenotype, followed by slow genetic assimilation of the new phenotype and reduction of plasticity. An association of colonization with increased plasticity depends on the difference in the optimal phenotype between ancestral and colonized environments, the difference in mean, variance and predictability of the environment, the cost of plasticity, and the time elapsed since colonization. The relative importance of these parameters depends on whether a phenotypic character develops by one-shot plasticity to a constant adult phenotype or by labile plasticity involving continuous and reversible development throughout adult life.