Data taken primarily from the Australian Youth Survey were used to model unemployment as a function of personal characteristics, family structure, and neighborhood composition using binomial probit estimation techniques. The cross-sectional model developed indicated that significant neighborhood effects on unemployment outcomes exist in high- and low-income areas. Although the positive effects of living in a high-income neighborhood diminished by age 21, the negative impacts associated with low-income neighborhoods persisted. Low neighborhood concentrations of vocational qualifications affected young people's unemployment outcomes. This finding was hypothesized to be an indicator of the weaker employment and information networks that typically exist in low-income neighborhoods. The panel data model indicated that unobserved heterogeneity (unobservable individual characteristics) influences the probability of being unemployed and is therefore another important factor in the modeling of neighborhood effects. The panel model confirmed the presence of neighborhood effects in the lowest 20% of neighborhoods but did not corroborate the existence of effects related to the concentration of vocational qualifications. A description of the variable set is appended. (Contains 8 tables and 35 references.) (MN)