Depression is a complex condition which changes over development. The focus of this PhD is on the developmental period from adolescence to emerging adulthood, but I also examined symptom development during a major population stressor in adulthood (i.e., COVID-19 pandemic) and during older adulthood. The overarching aim of this thesis was to understand the development of depressive symptoms from different perspectives. Across different developmental periods, I therefore aimed to (1) test measurement invariance (MI) of depression scales to ensure valid interpretation of the depression construct over time, (2) disentangle time-invariant and time-varying variance components (latent trait state occasion analysis) of the depression construct, (3) unravel symptom level associations and dynamics using network analysis, and (4) investigate differences in symptom reports of adolescents and their mothers. In Chapter 1, I outline complex developmental processes involved in depression. Chapter 2 addresses MI of depression scales from childhood to emerging adulthood. My systematic review of available evidence for gender, age, and longitudinal MI of child and adolescent depression scales revealed that MI is understudied in this age range. I then tested MI of the Short Mood and Feelings Questionnaire (SMFQ) in the Avon Longitudinal Study of Parents and Children (ALSPAC), a population-based study in South-West England (N = 7,364; ages 11-26). Longitudinal MI was not fully supported in models that included ages 11 and 13, but the highest level of MI (strict) was established from ages 14-26. At each wave, (partial) strict MI across gender was established. In Next Steps, a UK national cohort study (N = 13,539), MI of the General Health Questionnaire-12 (GHQ-12) was established from age 15 to 25 and across gender. Longitudinal and gender MI were also established for the eight-item Center for Epidemiological Studies Depression Scale (CES-D-8) in the English Longitudinal Study of Ageing (ELSA), a nine-wave representative study of the English population above 50 years of age (N = 11,391). These empirical results increase confidence in the interpretation of findings using total scores from SMFQ, GHQ-12, or CES-D-8 across these age ranges. Chapter 3 examines how much variance in symptoms of depression or psychological distress are attributable to stable time-invariant variance compared to transient time-varying variance. In ALSPAC, time-invariant variance in depressive symptoms increased from 8.1% at age 11 to 57.0% at age 26. Time-varying variance ranged from 91.9% at age 11 to 36.4% at age 26. Small proportions of the variance were explained by autoregressive pathways. Maternal depression, relationship with parents, stressful life events, school enjoyment, and previous mental and general health problems predicted time-varying variance at different ages. In Understanding Society, a UK probability- based longitudinal study of adults, using pre-pandemic (2015-2020) and pandemic data (N = 17,761, April 2020 to March 2021), time-varying variance in psychological distress was higher at the first COVID-19 lockdown (April 2020, 61.2%) compared to before the pandemic (~50%), suggesting increased fluctuations in distress at the start of the pandemic. Loneliness most strongly predicted time-varying variance during the first lockdown. These studies highlight the presence of fluctuations in mental health symptoms during adolescence and a major population stressor, but also indicate time-sensitive periods for intervention. In Chapter 4, I apply network modelling to understand dynamics of depressive symptoms across development beyond a common pathogenic pathway perspective. Across adolescent and emerging adult population-based samples, central symptoms encompassed negative self-evaluations and self-worth, both cross-sectionally (Sample 1: the Mental Health of Children and Young People in Great Britain 1999 survey [N = 4,235, ages 11-15]; Sample 2: the Millennium Cohort Study in the UK 2015 [N = 11,176, age 14]; Sample 3: the mental health of young people looked after by local authorities in England 2002 survey [N = 643, ages 11-17] and over time [ALSPAC]). In looked-after children, I was a bad person constituted a central symptom, while this was among the least central symptoms in the other two cross-sectional datasets. In an older representative population (ELSA), everything was an effort, could not get going and loneliness emerged as consistent central symptoms across time. Nosological core symptoms such as not enjoying anything were not central in the examined data sets. These findings point to clinically relevant symptoms in the development and persistence of depression, which may also challenge nosological views of depression. Chapter 5 focuses on informant discrepancies between mothers and their children in ALSPAC between the ages of 10-17. Latent-Trait-Occasion analysis indicated that mothers conceptualized their children’s depressive symptoms as more stable (time-invariant) compared with adolescents who conceptualized their own symptoms as more fluctuating (time-varying). Response surface analyses indicated that especially incongruent ratings (adolescent rating > mother rating) led to increased subsequent depression in adolescents. These studies highlight the importance of parental educational efforts to increase awareness and monitoring of children’s depressive symptoms in adolescence. Chapter 6 synthesizes the results by discussing differences and similarities across the examined developmental periods, what the different statistical approaches offer, as well as nosological considerations and resulting research questions arising from the findings. Overall, this work has significant implications for prevention and intervention initiatives in depression in sensitive developmental periods.