This issue of The CBHSQ Report examines sources of misused prescription pain relievers in the United States overall, by demographic subgroups, and by type of user. When comparing across subgroups of three or more levels, log-linear chi-square tests of independence of the subgroups and the outcome of interest were conducted in order to control the error level for multiple comparisons. When Shah’s Wald F test indicates overall significant differences, individual pairwise comparisons between subgroup levels (e.g., one subgroup level vs. another subgroup level) were conducted within SUDAAN to properly account for the complex sample design. Only those pairwise comparisons found to be significant at the .05 level are discussed in this report. All findings presented in this report are annual averages based on combined 2013 and 2014 NSDUH data.