The estimate of future obsolete streams is one of the crucial issues for the establishment of an efficient waste collection and recycling system in China. Due to low availability of reliable data, information on discarded household appliances (HAs) is deficient in China. This study adopts a stocks-based prediction model based on material flow analysis. The model firstly models the lifetime distribution of HAs, and then the future stocks of HAs are extrapolated. By determining the initial year of calculation, the model makes a prediction of future obsolete HAs in China in the time period from 2010–2030. The results show that the discarded amount of the five major kinds of HAs will increase from 130 million units in 2010 to 216 to 221 million units by 2020, and 259 to 282 million units by 2030. A total of 4370 to 4528 million units (149 to 155 million tonnes) of obsolete HAs will be generated in China over the next 20 years. Urban households will generate significantly more obsolete HAs (about 2619 to 2723 million units) than rural households, mainly due to the difference in their HAs possession levels. Thus recycling capacity must increase if the rising quantity of domestic obsolete HAs is to be handled properly. The results of this study can help to develop the collection and recycling systems and facilities needed for the obsolete HAs generated in the future. From a methodological perspective, the stock-based model provides a suitable tool to predict the generation of discarded HAs in the future