Objective: This study focused on interrelationships among supervisor attachment, supervisory working alliance, and supervision-related affect, plus the moderating effect of a field instructor training. Method: The researchers employed a pretest-posttest follow-up design of 100 randomly assigned field instructors and 64 students in two universities, using linear growth models to estimate relations among variables concurrently and over time. Result: Positive associations existed between attachment anxiety and negative affect among field instructors and between positive affect and supervisory alliance for all participants. Results suggested the training might have allowed field instructors to separate initial impressions of the alliance from change in positive affect over the year. Conclusion: This study underscores the need for research on field instructor training models and attachment styles in supervision. (Contains 3 tables.)