This paper evaluates the impact of 33/50 program on 33/50 releases and examine if this impact differed for ozone depleting chemicals and the rest 15 chemicals. It also examines the extent to which any reduction in 33/50 releases can be attributed to pollution prevention techniques (P2). Our analysis is conducted at the facility level, using facility level participation information and emissions data for the 1988-1995 periods. The data set includes 12,463 facilities eligible to participate in the program starting in 1991. Of these, 1033 facilities belonging to 1268 parent companies participated in the 33/50 program. Dynamic panel data models are used to incorporate facility specific unobserved effect, timing of participation and endogeneity of program participation and P2 adoption decisions. The results show that the program has different impact on ozone depleting chemicals and other releases. Regulatory pressure to comply with Montreal protocol motivated firms to reduce emissions during early stage of the program. When the immediate threat or regulation was reduced, participants appear to have focused on non-ozone depleting chemicals. The pollution prevention methods adopted also contributed to a reduction in toxic releases.