Coaching is a systematic and goal-oriented one-on-one intervention by a coach aimed to guide clients in their professional and personal development. Previous research on coaching has demonstrated effects on a number of positive outcomes, including well-being and performance, yet little is known about the processes that underlie these outcomes, such as the type of questions coaches use. Here, we focus on three different types of coaching questions, and aim to uncover their immediate and sustained effects for affect, self-efficacy, and goal-directed outcomes, using a between-subjects experiment. One hundred and eighty-three medical residents and PhD students from various medical centers and healthcare organizations in the Netherlands were recruited to participate in a self-coaching writing exercise, where they followed written instructions rather than interacting with a real coach. All participants were randomly allocated to one of three conditions: either one of two solution-focused coaching conditions (i.e., the success or miracle condition) or a problem-focused coaching condition. Self-report questionnaires were used to measure key outcomes of coaching, that is positive and negative affect, self-efficacy, goal orientation, action planning (i.e., quantity and quality) and goal attainment. Two follow-up measurements assessed if the effects of the self-coaching exercise led to problem-solving actions within an initial follow-up period of 14 days and a subsequent follow-up period of 10 days. Findings showed that participants experienced more positive affect, less negative affect, and higher approach goal orientation after the solution-focused coaching exercise compared to the problem-focused coaching exercise. In all conditions, goal attainment increased as a consequence of the self-coaching intervention. We discuss the implications of our findings for the science and practice of contemporary coaching.