A growing body of research (Blumberg & Ong, 1998; Danziger et al., 2000) documents lack of access to reliable transportation as a major barrier to employment, and a positive relationship between car ownership and employment in both rural and urban areas (Fletcher, Garasky, & Jensen, 2002; Ong, 2002). Despite the growing evidence that reliable transportation is a critical component of public programs that promote self-sufficiency, many public assistance programs incorporate vehicle asset limits in a means test of eligibility. While discussing vehicle ownership among Temporary Assistance for Needy Families (TANF) recipients, Ong (2002) described these policies as ". . .still shaped by an earlier and largely unfounded fear that welfare recipients will waste resources on luxury cars" (p.249). Using data from the Iowa Transportation and Employment Survey (ITES), we explore the relationship between vehicle value and vehicle reliability, paying particular attention to the vehicle asset limits of Iowa's TANF program. Each of the 768 respondents in the ITES provided a complete roster of all vehicles in the household, including the year, make and model and an assessment of reliability using a 5-point scale ranging from "very reliable" to "not working at all." We assigned a "trade-in" value from the National Automobile Dealer's Association (NADA) Used Car Guide Book for each vehicle. Vehicles were excluded from the database if incomplete information made it impossible to identify a specific vehicle model and, therefore, its estimated value. This process resulted in a sample of 1,249 vehicles used in the analysis. Contrary to stereotypic views that are portrayed in the press, our analyses suggest that TANF recipients (and other low-income households) do not own "luxury" cars. Only I % of the recipient-owned vehicles were valued at more than $7,550 in 2001. Descriptive analysis shows that the proportion of vehicles that are evaluated as "very reliable" is much higher among vehicles in the highest quintile of the vehicle distribution arrayed by estimated value. About 42% versus 96% of the vehicles in the lowest and highest quintiles are rated "very reliable." Further, our analysis using a regression simulation suggests that vehicles with asset values that exceed Iowa's TANF vehicle asset limit ($3,889) have a higher probability of being "very reliable" compared to vehicles with values below this cutoff. Our findings suggest that public policy and programs that facilitate the purchase and ownership of more reliable vehicles would, in turn, support the employment goals of TANF. [ABSTRACT FROM AUTHOR]