Polyglutamine protein aggregates are a hallmark of several neurodegenerative diseases, including Huntington's disease, and increasing evidence suggests that reducing or inhibiting aggregation produces a therapeutic benefit in animal models of disease. Part of the challenge in designing compounds that interfere with protein aggregation is having a sensitive and consistent in vitro assay that allows for efficient screening and lead optimization. Here we describe a simplified polyglutamine assay that uses a soluble, pathological-length polyglutamine construct (62 glutamines [Q62]) fused to glutathione-S-transferase (GST) and measure aggregate formation with fluorescence generated by thioflavin T binding. Controlled release of Q62 from GST using proteolytic cleavage resulted in time-dependent aggregate formation that was not observed for a non-pathological-length GST-Q19 construct. Cleavage of the polyglutamine domain from GST increased the rate of Q62 aggregation from days to hours, significantly decreasing the time for compound analysis. Controlled aggregate formation combined with the fluorescence sensitivity of the dye thioflavin T allowed us to screen a series of peptide analogs for lead optimization of a previously identified peptide aggregation inhibitor, QBP1. QBP1 analogs showed the greatest inhibitory potency when added prior to Q62 aggregate initiation, suggesting that the mechanism of inhibition was interference with early formed aggregates that were not detectable by ultraviolet or dye binding. The assay detected activities that differed by three orders of magnitudes with Z' = 0.56, which is suitable for high-throughput screening and allowed us to do lead optimization of QBP1 analogs for pharmacophore model building.