The circumplex model of affect posits that valence and arousal are the principal dimensions of affect. The center of the 2D space represents a neutral state of valence and a medium state of arousal. The role of valence and arousal in human emotion has been studied extensively. However, no consistent relationship between valence and arousal has been established. Most of the prior studies investigating the relationship have been conducted in relatively controlled laboratory settings. Ecological momentary assessment (EMA) of affect from participants residing in permanent supportive housing was used to study the relationship between valence and arousal in real-life settings. The goal of this study was to explore the relationship between valence and arousal in a person's natural environment. Participants were recruited from housing agencies in Fort Worth, Texas, United States. All participants had a history of chronic homelessness and reported at least one mental health condition. A subset of participants completed daily (morning) EMAs of emotions and other behaviors. The sample comprised 78 women and 77 men, and the average age was 52 (SD 8) years. From the circumplex model of affect, the EMA included 9 questions related to the participant's current emotional state (happy, frustrated, sad, worried, restless, excited, calm, bored, and sluggish). The responses were used to calculate 2 composite scores for valence and arousal. Statistical models uniformly showed a dominant linear relation between valence and arousal and a significant difference in the slopes among races. None of the other effects were statistically significant. Compared with previous studies, the effects were quite robust. Our findings may provide a window to the fundamental structure of affect. We found a strong positive linear relationship between valence and arousal at the nomothetic level, which may provide insight into a universal structure of affect. However, the study needs to be replicated for different populations to determine whether our findings can be generalized beyond the population studied here.