This paper reports on a study of 35 Mandarin Chinese learners who (1) created pitch curves of their spoken word tones and (2) compared their pitch curves with those of native speakers while practicing pronunciation. Following a pretest, the learners received training for 20–25 minutes weekly over nine weeks and took a posttest. Two types of data analyses were performed. First, native speakers of Mandarin auditorily rated the pretests and posttests. The ratings revealed that learners’ pronunciation of tones improved between pretest and posttest. Second, acoustic analyses of the learners’ recordings were conducted, and the learners’ production was compared with that of native speakers. Results indicated that students’ pronunciation of some tones improved in the posttest. The postsurveys indicated that two-thirds of the participants found viewing pitch curves helpful. This study confirms previous research but suggests that acoustic analyses complement auditory analyses with more precise indications of L2 learners’ tonal difficulties.