Phenotypic plasticity describes the ability of an individual to alter its phenotype in response to the environment and is potentially adaptive when dealing with environmental variation. However, robustness in the face of a changing environment may often be beneficial for traits that are tightly linked to fitness. We hypothesized that robustness of some traits may depend on specific patterns of plasticity within and among other traits. We used a reaction norm approach to study robustness and phenotypic plasticity of three life history traits of the collembolan Orchesella cincta in environments with different thermal regimes. We measured adult mass, age at maturity and growth rate of males and females from heath and forest habitats at two temperatures (12 and 22 °C). We found evidence for ecotype-specific robustness of female adult mass to temperature, with a higher level of robustness in the heath ecotype. This robustness is facilitated by plastic adjustments of growth rate and age at maturity. Furthermore, female fecundity is strongly influenced by female adult mass, explaining the importance of realizing a high mass across temperatures for females. These findings indicate that different predicted outcomes of life history theory can be combined within one species’ ontogeny and that models describing life history strategies should not assume that traits like growth rate are maximized under all conditions. On a methodological note, we report a systematic inflation of variation when standard deviations and correlation coefficients are calculated from family means as opposed to individual data within a family structure.,Springtails measured traitsCollected data on offspring from field collected spingtails, at two temperatures. Column headings: familie = family of sibs from same egg batch, location = origin of original population, either Kampina nature reserve (k) or Hilversumse heide nature reserve (h), ecotype = forest (f) or heath (h), temp = temperature at which eggs were placed (12 or 22 C), time = age at maturity in days, sex = sex of individual (m= male or f= female), mass = mass at maturity in ug, rate = growth rate ug/days, lnmass/lnrate/lntime = natural logarithm of the traits mass, rate and time as used in random regression model.Springtails.csvfecundityfecundity of female springtails measured as number of eggs in first egg batch. Column headings: familie = family of sibs from same egg batch, location = origin of original population, either Kampina nature reserve (k) or Hilversumse heide nature reserve (h), ecotype = forest (f) or heath (h), temp = temperature at which mother developed (12 or 22 C), eggnumber = number of eggs in first laid batch, mass_mother = mass of the mother when first egg batch was produced, lnmass_mother = natural logarithm of mass_mother used in linear mixed model.eggvolumeeggvolume of eggs used to set up experiment. Column headings: familie = family of sibs from same egg batch, location = origin of original population, either Kampina nature reserve (k) or Hilversumse heide nature reserve (h), ecotype = forest (f) or heath (h), volmm3 = volume of egg in mm3.,