The Toxics Exposure Assessment Columbia-Harvard (TEACH) Project characterized personal, indoor, and outdoor concentrations of a suite of volatile organic compounds (VOCs) for high school students living in New York City (NYC) and Los Angeles (LA). This paper presents the analysis of VOC measurements collected indoors and outdoors for 46 students’ homes in NYC and for 41 students’ homes in LA across two seasons. Dual-sorbent thermal desorption tubes were used for the collection of 15 VOCs and C18 2,4-dinitrophenylhydrazine-coated cartridges were used for the collection of seven aldehydes. Air-exchange rates (AERs) were also measured using a perfluorocarbon tracer gas method. The AERs were lower in the winter in both cities, averaging 1 h-1 in NYC and 1.4 h-1 in LA, compared with 1.8 h-1 in NYC in the summer and 2.5 h-1 in LA in the fall. Higher AERs were generally associated with lower indoor–outdoor ratios with significant differences for the compounds with indoor sources, including chloroform, 1,4-dichlorobenzene, and formaldehyde. Using a mass-balance model to account for AER and other housing parameters, effective source emission rates (SER) were calculated for each compound. Based on I/O ratios and source emission rates, VOCs could be divided into: (1) indoor-source-influenced compounds, (2) those with contributions from both indoor and outdoor sources, and (3) those with mostly outdoor sources. Significant indoor sources were found for the following six compounds (mean emission rates presented): chloroform (0.11 mg/h), 1,4-dichlorobenzene (19 mg/h), formaldehyde (5 mg/h), acetaldehyde (2 mg/h), benzaldehyde (0.6 mg/h), and hexaldehyde (2 mg/h). Although chloroform had variable I/O ratios across seasons, SERs, which accounted for AER, were similar in both cities for both seasons (e.g., LA means 0.12 and 0.11 mg/h in winter and fall, respectively). Formaldehyde had substantially higher indoor emission rates in the summer in NYC compared to winter (3.8 vs. 1.6 mg/h) but lower in the fall in LA compared to winter (4.3 vs. 5.0 mg/h). Uncertainty analysis determined that source strength calculations were not sensitive to measurement error for a subset of homes in LA.