This article reviews the arguments that support the approach regarding the 'inflationary' influence of the negative affectivity (NA) trait, or neuroticism, on relationships between perceived stress (e.g. different means of self-report of psychosocial stress such as major life events or daily stress) and subjective physical health (e.g. somatic symptomatology, health complaints). It has been explained that the said inflationary effect could distort the interpretation of the results that are found if it is done in terms of stress-objective health relationships (known in terms of health behavior, biological markers, morbidity and mortality) and that, in this sense, the NA would be a contaminant. This is why it has been suggested that in these cases, the effect of NA on predictor and criterion relationships should be controlled. The present critical review doesn't support these conclusions. Also, this paper deals with the empirical analysis of some matters that refer to one of the arguments mentioned, namely that NA is a general and penetrating factor that impregnates perceived stress (we will focus on daily stress or hassles) and subjective health measures (we will focus on somatic symptoms), as well as its relationships. We also analyze the role of positive affectivity (PA). The empirical analysis carried out uses a combination of within and interindividual methods. Contrary to the hypothesis regarding the moderating effects (specifically, inflationary) of NA on relationships between daily stress and somatic symptomatology, NA has only presented direct significant effects on both variables (between subjects), while its moderating effects are modest and non-significant. However, PA appears not to have any direct significant effects, but it does have moderating effects (specifically, buffering effects). Finally, we discuss the implications of these findings and we make some suggestions for future research.