A joint Norwegian–Greenland longline survey was conducted at East Greenland in August 1997, using different hook and bait types. Most Norwegian longliners use hooks of the type EZ 12/0. This hook was compared to three versions of a new circle 14/0 hook. A total catch of 2899 Greenland halibut from 45,760 hooks baited with squid were used in the hook selectivity analyses. In average, CPUE was 281kg/1000 hook for the EZ hook. CPUE for the circular hook was 36% higher making an overall significant difference in CPUE between the EZ hook and the circle hooks. On 6630 hooks squid and grenadier were used alternately. The CPUE of Greenland halibut was 25% higher for grenadier bait. The grenadier bait resulted in a reduction in bycatch compared to the squid bait (1.1 and 20.7% by numbers, respectively). Catches by EZ 12/0 hook and one of the circle 14/0 hooks were compared in order to examine size selectivity. Using the SELECT approach, expected proportions were fitted to the observed proportions for five different models of selectivity. All models resulted in almost identical fits. The absence of non-selective data requires the choice of selectivity curve to be based on knowledge about the capture process. Since the selectivity curves cannot be determined unambiguously in this study, none of the estimated curves are preferred for the other.