Objective To analyze the age-related changes in muscle mass within the study population aged 6 to 70 years in Guangzhou， and to explore the muscle development patterns among individuals with different weight statuses. Methods A cross-sectional study was carried out， encompassing 3，316 residents aged 6–70 years in Guangzhou. Dual-energy X-ray absorptiometry was employed to measure muscle mass. Subsequently， the skeletal muscle mass index （SMMI） and appendicular skeletal muscle mass index （ASMMI） were computed after adjusting for body height. The LMS method was utilized to fit the sex-and age-specific percentile curves of muscle indices， and a stratified analysis was performed according to body weight status. Results Muscle mass in males was greater than that in females across all age groups. Both genders showed rapid SMMI growth between the ages of 6 and 12 years， with average annual growth rates of 4.6% for males and 3.9% for females. For males， the growth rate of SMMI slowed down after the age of 12 years， with an average annual growth of 1.9% from 13 to 19 years， and then entered a slow-growth phase after the age of 20 years， reaching its peak at the age of 43 years. For females， SMMI increased by an average of 0.8% per year from 13 to 19 years， reached its peak at the age of 47 years， and further decreased after the age of 55 years. The age-related changes in ASMMI were similar to those of SMMI， although the peak ASMMI occurred at an earlier age. When stratified by body weight status， individuals with obesity had the highest peak SMMI but the most rapid decline after the peak age. In contrast， underweight individuals had the lowest muscle mass peak but a much slower decline trend. Conclusions Results indicate a distinct age- and sex-specific pattern of muscle mass development within the study population. The decline of muscle mass during middle and old age exhibits site-specific characteristics， with appendicular skeletal muscle experiencing the earliest decline. The changes in muscle mass among obese individuals follow a pattern of “high peak value – rapid decline”. Owing to the non-random sampling design， the generalizability of the findings is restricted. Nevertheless， the findings can offer preliminary evidence for promoting muscle health.