As part of the Baltimore PM2.5 Supersite study, intensive three-hourly continuous PM2.5 sampling was conducted for nearly 4 weeks in summer of 2002 and as well in winter of 2002/2003. Close to 120 individual organic compounds have been quantified separately in filter and polyurethane foam (PUF) plug pairs for 17 days for each sampling period. Here, the focus is on (1) describing briefly the new sampling system, (2) discussing filter/PUF plugs breakthrough experiments for semi-volatile compounds, (3) providing insight into phase distribution of semi-volatile organic species, and (4) discussing the impact of air pollution sampling time on human exposure with information on maximum 3- and 24-h averaged ambient concentrations of potentially adverse health effects causing organic pollutants. The newly developed sampling system consisted of five electronically controlled parallel sampling channels that are operated in a sequential mode. Semi-volatile breakthrough experiments were conducted in three separate experiments over 3, 4, and 5 h each using one filter and three PUF plugs. Valuable insight was obtained about the transfer of semi-volatile organic compounds through the sequence of PUF plugs and a cut-off could be defined for complete sampling of semi-volatile compounds on only one filter/PUF plug pair, i.e., the setup finally used during the seasonal PM2.5 sampling campaign. Accordingly, n-nonadecane (C19) with a vapor pressure (vp) of 3.25 × 10(-4) Torr is collected with > 95% on the filter/PUF pair. Applied to phenanthrene, the most abundant the PAH sampled, phenanthrene (vp, 6.2 × 10(-5) Torr) was collected completely in wintertime and correlates very well with three-hourly PM2.5 ambient concentrations. Valuable data on the fractional partitioning for semi-volatile organics as a function of season is provided here and can be used to differentiate the human uptake of an organic pollutant of interest via gas- and particle-phase exposure. Health effects studies often relay on PM2.5 exposure measurements taken over 24 h or longer. We found that maximum 3-h concentrations are frequently two to five times higher than that found for maximum 24-h concentrations, an important aspect when considering that short-term exposure to higher air pollution levels are more likely to overpower defense mechanisms in the human lung with subsequent adverse effects even at lower pollutant levels.