A new source-type identification method, Reduction and Species Clustering Using Episodes (ReSCUE), was developed to exploit the temporal synchronicity typically observed between ambient species in high time resolution fine particulate matter (PM2.5) data to form clusters that vary together. High time-resolution (30 min) PM2.5 sampling was conducted for a month during the summer of 2006 in Steubenville, OH, an EPA designated nonattainment area for the U.S. National Ambient Air Quality Standards (NAAQS). When the data were evaluated, the species clusters from ReSCUE matched extremely well with the source types identified by EPA Unmix demonstrating that ReSCUE is a valuable tool in identifying source types. Results from EPA Unmix show that contributions to PM2.5 are mostly from iron/steel manufacturing (36% ± 9%), crustal matter (33% ± 11%), and coal combustion (11% ± 19%). More importantly, ReSCUE was useful in (i) providing objective data driven guidance for the number of source factors and key fitting species for EPA Unmix, and (ii) detecting tenuous associations between some species and source types in the results derived by EPA Unmix.