This is the first systematic acoustic analysis of Putonghua tone errors produced by Hong Kong speakers. Putonghua tones produced in three contexts were investigated: isolated citation form, in the middle of a sentence, and at the end of a sentence. Acoustic data of f_0 (Hz) and duration (s) were measured and analyzed. The results showed that different contexts have an important effect on tone error patterns. There are more severe tone errors in the middle of a sentence, especially the pitch contours of Tone 2 and Tone 3 are very close in the middle of sentence, which seem very likely to be merged. The studies conducted here provided an accurate description of Putonghua tone errors made by Hong Kong speakers in different contexts, which resulted in reasonable suggestions for teaching and learning strategies for them.