This study reports the response of snapper Pagrus auratus to the establishment of no-take status in a marine reserve around the Poor Knights Islands in northeastern New Zealand. The Poor Knights and 2 reference locations, Cape Brett and the Mokohinau Islands, were sampled biannually for 4 yr using baited underwater video (BUV). Following the implementation of full marine reserve status at the Poor Knights in October 1998, snapper showed significant increases in abundance and biomass relative to fished control locations. This was particularly apparent for large snapper (>270 mm), whose numbers increased rapidly to levels 7.4 times higher in the final survey compared to the initial pre-reserve survey, and total snapper biomass increased by 818%. There was no significant increase in the abundance, biomass or size of snapper at the reference locations over this time. There was a strong seasonal trend in snapper abundance, with higher numbers in autumn (March/April) compared to spring (September/October). The daily batch fecundity was 11 to 18 times higher at the Poor Knights compared to the reference locations. Once fishing ceased in previously partially protected areas, a rapid recovery of snapper ensued, suggesting that partial fishing regulations are ineffective for protecting targeted species. The speed of increase in snapper density resulted from the immigration of adult fish into the reserve, rather than from within-reserve recruitment.