This qualitative study examines communication strategies (CS) for linguistic problems in second language (L2) oral communication, addressing two research questions: (a) What types of CS are used to manage linguistic problems in L2 oral communication? and (b) What factors affect the choice between abandon and achievement CS? Data in the form of interviews, observations, conversations, stimulated recalls on the recorded conversations, and measurements of language proficiency and personal characteristics were collected from 12 Korean students studying in a university in the U.S. From the inductive and taxonomic analyses, seven types of CS were identified: (a) phonetic CS, (b) lexical CS, (c) syntactic CS, (d) dual lexical-syntactic CS, (e) semantic CS, (f) time-gaining CS, and (g) comprehension CS, each of which were further classified into subcategories in multiple levels. The inductive and negative case analyses also revealed that the choice between abandon CS and achievement CS was affected by various antecedent factors, which interacted and were organized in multiple levels. The choice of CS was determined by the following procedural steps. First, various personal, environmental, andpersonal-environmental factors jointly influenced the participant's perceived psychological antecedent conditions: (a) security, (b) excitement, (c) responsibility, and (d) manageability. Second, these four perceived psychological conditions jointly constructed the participant's willingness to solve the problem, which was the ultimate factor affecting the choice of CS. Third, the influence of willingness to solve the problem on the choice was mediated by two other factors: (a) the estimated cognitive load entailed in the use of CS, which was controlled by the complexity of the problem and of the anticipated CS and (b) self-empowerment for communicative achievement,which was constrained by an individual's self-expressiveness and self-comfort. The individual's self-expressiveness, which is relatively stable, was also subject to the individual's characteristics, such as gender, shyness, language proficiency, self-perceived communicative competence, communication apprehension, and identities. The individual's self-comfort easily fluctuated in accordance with the individual's physical and psychological conditions. These findings demonstrate the usefulness of the ecological perspective in examining communication behaviors, including the choice between abandon CS and achievement CS and the need for individualized, rather than uniform, pedagogical intervention.