As personalized treatment options gain popularity in healthcare and evidence of the heterogeneity of mental health trajectories grows, clinical psychology practice can improve patient outcomes by leveraging these individual trajectories. This research investigates whether mental health trajectories are predictable in individuals, specifically whether time-series forecasting tools can generate accurate forecasts of single-subject clinical data. Moreover, we investigate whether an increasing amount of data improves forecasting accuracy, revealing whether practitioners can reliably employ these tools to make data-driven decisions. To test these questions, nine simulated N=1 time series and an empirical N=1 time-series dataset of depressive symptom severity are used to assess the predictive accuracy of various forecasting models for clinical symptoms and investigate the relationship between forecasting accuracy and the amount of training data. Findings may promote the use and development of forecasting tools in clinical practice while encouraging a shift towards more personalized and effective mental health care.