In 2010, only 19 percent of individuals ages 50-58 whose household incomes were less than 300 percent of the poverty line participated in a pension of any kind at their current jobs, compared to 56 percent of those above 300 percent of poverty.  This paper investigates this pension gap.  In particular, we decompose the pension participation rate into its four elements in order to compare coverage between higher- and lower-income individuals: 1) the fraction of people who are currently working (the employment rate); 2) the fraction of workers who are in firms that offer pension benefits to at least some workers (the offer rate); 3) the fraction of workers who are eligible for pension benefits, conditional on being in a firm where it is offered (the eligibility rate); and 4) the fraction of workers who enroll in a pension plan when they are eligible (the take-up rate).  We find that the substantial pension gap between higher- and lower-income individuals is driven primarily by the lower-income group’s lower employment rate and the smaller probability of working for an employer that offers pensions; when lower-income workers do have a pension plan at work, their eligibility and take-up rates are nearly equivalent to higher-income workers.  We also find that the factors associated with a higher value for each element of pension participation are very consistent: higher education and income, previous pension history, and job characteristics including firm size, occupation, job tenure, and union status.  Together, these findings suggest that policies such as automatic enrollment that focus on pension eligibility or take-up are unlikely to close the pension coverage gap between older, lower-income individuals and their higher-income contemporaries; instead, greater pension participation requires more jobs and, in particular, more “good jobs.”