The dimeric β-barrel protein Cu, Zn superoxide dismutase (SOD1) is an optimal model system for investigating the molecular basis for toxicity in ALS. Mutations throughout the β-barrel sequence are causative for ALS and implicated in the gain-of-function toxicity. The occurrence of mutations at dozens of sites suggests a common mechanism of toxicity for SOD1 disease causing variants. Previous biophysical and structural data on SOD1 and its variants have established a fairly detailed understanding of its folding energy landscape. A key finding is that ALS-variants of SOD1 lead to enhanced populations of both folded and unfolded monomeric forms of SOD1. An unresolved question, however, is the mechanism of oligomerization and toxicity of SOD1 variants. To address this question, we present an aggregation assay that matches physiological conditions as much as possible, avoiding denaturants, fluorinated alcohols and agitation. Our results, based on the analysis of the WT protein and several variants, suggest that the oligomerization rate is correlated with the population of the unfolded monomer. Furthermore, although a secondary nucleation mechanism has been proposed as a key step in SOD1 aggregation our results, obtained without agitation, are best fitted with a crystallization-like nucleation model without fragmentation. We test the consistency of in vitro and live cell data by performing fluorescence correlation spectroscopy (FCS) measurements on live HEK293 cells expressing EGFP fusion constructs of WT and disease causing SOD1 variants. Preliminary quantitative modeling of the FCS data suggest consistency in the propagation rates between in vitro and live cell data. The implication of the unfolded monomer of SOD1 suggests that small-molecules stabilizing the dimeric native state can be used as a potential therapeutic strategy.This research was supported by NIH grant GM54836.