BACKGROUND:
Internships represent a pivotal school-to-work transition during which turnover intention may emerge and shape subsequent career adjustment. However, evidence on heterogeneous longitudinal patterns of turnover intention and their implications for later mental health and occupational well-being remains limited.

METHODS:
Three-wave longitudinal data were collected from 352 interns in a large automobile manufacturing enterprise in China (March 2024, November 2024, and April 2025). Latent class growth analysis (LCGA) was conducted to identify distinct trajectories of turnover intention across Time 1 (T1), Time 2 (T2), and Time 3 (T3). Baseline predictors of class membership were examined using a three-step multinomial logistic regression procedure (R3STEP). Distal outcomes at T3 were compared using the Bolck–Croon–Hagenaars (BCH) method, including depression, anxiety, job satisfaction, and mature happiness.

RESULTS:
A two-class solution was selected (entropy = 0.755). Class 1 showed a low-initial, slightly increasing trajectory (intercept = 1.419; slope = 0.201), whereas Class 2 showed a high-initial, decreasing trajectory (intercept = 2.622; slope = −0.445). Higher baseline work stress and a higher perceived extent of machine use in the workplace increased the likelihood of belonging to Class 2, whereas higher person–job fit decreased this likelihood; workplace support was positively associated with Class 2 after covariate adjustment. At T3, compared with Class 1, Class 2 reported higher depression (M = 1.760 vs. 1.231) and anxiety (M = 1.630 vs. 1.144), and lower job satisfaction (M = 3.518 vs. 3.978) and mature happiness (M = 3.131 vs. 3.849).

CONCLUSIONS:
Interns’ turnover intention is heterogeneous over time and is associated with subsequent mental health and occupational well-being. Early-career interventions should prioritize stress management, strengthening person–job fit, and supporting adaptation in digitally intensive workplaces.